conference
stringclasses
6 values
title
stringlengths
2
176
abstract
stringlengths
2
5k
decision
stringclasses
11 values
ICLR.cc/2022/Conference
On the Importance of Difficulty Calibration in Membership Inference Attacks
The vulnerability of machine learning models to membership inference attacks has received much attention in recent years. However, existing attacks mostly remain impractical due to having high false positive rates, where non-member samples are often erroneously predicted as members. This type of error makes the predicted membership signal unreliable, especially since most samples are non-members in real world applications. In this work, we argue that membership inference attacks can benefit drastically from difficulty calibration, where an attack's predicted membership score is adjusted to the difficulty of correctly classifying the target sample. We show that difficulty calibration can significantly reduce the false positive rate of a variety of existing attacks without a loss in accuracy.
Accept (Poster)
ICLR.cc/2021/Conference
Shapley Explanation Networks
Shapley values have become one of the most popular feature attribution explanation methods. However, most prior work has focused on post-hoc Shapley explanations, which can be computationally demanding due to its exponential time complexity and preclude model regularization based on Shapley explanations during training. Thus, we propose to incorporate Shapley values themselves as latent representations in deep models thereby making Shapley explanations first-class citizens in the modeling paradigm. This intrinsic explanation approach enables layer-wise explanations, explanation regularization of the model during training, and fast explanation computation at test time. We define the Shapley transform that transforms the input into a Shapley representation given a specific function. We operationalize the Shapley transform as a neural network module and construct both shallow and deep networks, called ShapNets, by composing Shapley modules. We prove that our Shallow ShapNets compute the exact Shapley values and our Deep ShapNets maintain the missingness and accuracy properties of Shapley values. We demonstrate on synthetic and real-world datasets that our ShapNets enable layer-wise Shapley explanations, novel Shapley regularizations during training, and fast computation while maintaining reasonable performance. Code is available at https://github.com/inouye-lab/ShapleyExplanationNetworks.
Accept (Poster)
ICLR.cc/2022/Conference
Neural Face Identification in a 2D Wireframe Projection of a Manifold Object
In computer-aided design (CAD) systems, 2D line drawings are commonly used to illustrate 3D object designs. To reconstruct the 3D models depicted by a single 2D line drawing, an important key is finding the edge loops in the line drawing which correspond to the actual faces of the 3D object. In this paper, we approach the classical problem of face identification from a novel data-driven point of view. We cast it as a sequence generation problem: starting from an arbitrary edge, we adopt a variant of the popular Transformer model to predict the edges associated with the same face in a natural order. This allows us to avoid searching the space of all possible edges loops with various hand-crafted rules and heuristics as most existing methods do, deal with challenging cases such as curved surfaces and nested edge loops, and leverage additional cues such as face types. We further discuss how possibly imperfect predictions can be used for 3D object reconstruction.
Withdrawn
ICLR.cc/2022/Conference
When less is more: Simplifying inputs aids neural network understanding
Are all bits useful? In this work, we propose SimpleBits, a method to synthesize simplified inputs by reducing information content, and carefully measure the effect of such simplification on learning. Crucially, SimpleBits does not require any domain-specific knowledge to constrain which input features should be removed. Instead, SimpleBits learns to remove the features of inputs which are least relevant for a given task. Concretely, we jointly optimize for input simplification by reducing inputs' bits per dimension as given by a pretrained generative model, as well as for the classification performance. We apply the simplification approach to a wide range of scenarios: conventional training, dataset condensation and post-hoc explanations. In this way, we analyze what simplified inputs tell us about the decisions made by classification networks. We show that our simplification approach successfully removes superfluous information for tasks with injected distractors. When applied post-hoc, our approach provides intuition into reasons for misclassifications of conventionally trained classifiers. Finally, for dataset condensation, we find that inputs can be simplified with only minimal accuracy degradation. Overall, our learning-based simplification approach offers a valuable new tool to explore the basis of network decisions.
Reject
ICLR.cc/2019/Conference
Beyond Greedy Ranking: Slate Optimization via List-CVAE
The conventional approach to solving the recommendation problem greedily ranks individual document candidates by prediction scores. However, this method fails to optimize the slate as a whole, and hence, often struggles to capture biases caused by the page layout and document interdepedencies. The slate recommendation problem aims to directly find the optimally ordered subset of documents (i.e. slates) that best serve users’ interests. Solving this problem is hard due to the combinatorial explosion of document candidates and their display positions on the page. Therefore we propose a paradigm shift from the traditional viewpoint of solving a ranking problem to a direct slate generation framework. In this paper, we introduce List Conditional Variational Auto-Encoders (ListCVAE), which learn the joint distribution of documents on the slate conditioned on user responses, and directly generate full slates. Experiments on simulated and real-world data show that List-CVAE outperforms greedy ranking methods consistently on various scales of documents corpora.
Accept (Poster)
ICLR.cc/2023/Conference
How does Uncertainty-aware Sample-selection Help Decision against Action Noise?
Learning from imperfect demonstrations has become a vital problem in imitation learning (IL). Since the assumption of the collected demonstrations are optimal cannot always hold in real-world tasks, many previous works considers learning from a mixture of optimal and sub-optimal demonstrations. On the other hand, video records can be hands-down demonstrations in practice. Leveraging such demonstrations requires labors to output action for each frame. However, action noise always occurs when the labors are not domain experts, or meet confusing state frames. Previous IL methods can be vulnerable to such demonstrations with state-dependent action noise. To tackle this problem, we propose a robust learning paradigm called USN, which bridges Uncertainty-aware Sample-selection with Negative learning. First, IL model feeds forward all demonstration data and estimates its predictive uncertainty. Then, we select large-loss samples in the light of the uncertainty measures. Next, we update the model parameters with additional negative learning on the selected samples. Empirical results on Box2D tasks and Atari games demonstrate that USN improves the performance of state-of-the-art IL methods by more than 10% under a large portion of action noise.
Reject
ICLR.cc/2022/Conference
Modelling neuronal behaviour with time series regression: Recurrent Neural Networks on synthetic C. elegans data
Given the inner complexity of the human nervous system, insight into the dynamics of brain activity can be gained from understanding smaller and simpler organisms, such as the nematode C. elegans. The behavioural and structural biology of these organisms is well-known, making them prime candidates for benchmarking modelling and simulation techniques. In these complex neuronal collections, classical white-box modelling techniques based on intrinsic structural or behavioural information are either unable to capture the profound nonlinearities of the neuronal response to different stimuli or generate extremely complex models, which are computationally intractable. In this paper we investigate whether it is possible to generate lower complexity black-box models that can capture the system dynamics with low error using only measured or simulated input-output information. We show how the nervous system of C. elegans can be modelled and simulated with data-driven models using different neural network architectures. Specifically, we target the use of state of the art recurrent neural networks architectures such as LSTMs and GRUs and compare these architectures in terms of their properties and their RMSE, as well as the complexity of the resulting models. We show that GRU models with a hidden layer size of 4 units are able to accurately reproduce the system's response to very different stimuli.
Reject
ICLR.cc/2023/Conference
Spotlight: Mobile UI Understanding using Vision-Language Models with a Focus
Mobile UI understanding is important for enabling various interaction tasks such as UI automation and accessibility. Previous mobile UI modeling often depends on the view hierarchy information of a screen, which directly provides the structural data of the UI, with the hope to bypass challenging tasks of visual modeling from screen pixels. However, view hierarchies are not always available, and are often corrupted with missing object descriptions or misaligned structure information. As a result, despite the use of view hierarchies could offer short-term gains, it may ultimately hinder the applicability and performance of the model. In this paper, we propose Spotlight, a vision-only approach for mobile UI understanding. Specifically, we enhance a vision-language model that only takes the screenshot of the UI and a region of interest on the screen---the focus---as the input. This general architecture of Spotlight is easily scalable and capable of performing a range of UI modeling tasks. Our experiments show that our model establishes SoTA results on several representative UI tasks and outperforms previous methods that use both screenshots and view hierarchies as inputs. Furthermore, we explore multi-task learning and few-shot prompting capacities of the proposed models, demonstrating promising results in the multi-task learning direction.
Accept: poster
ICLR.cc/2021/Conference
Bigeminal Priors Variational Auto-encoder
Variational auto-encoders (VAEs) are an influential and generally-used class of likelihood-based generative models in unsupervised learning. The likelihood-based generative models have been reported to be highly robust to the out-of-distribution (OOD) inputs and can be a detector by assuming that the model assigns higher likelihoods to the samples from the in-distribution (ID) dataset than an OOD dataset. However, recent works reported a phenomenon that VAE recognizes some OOD samples as ID by assigning a higher likelihood to the OOD inputs compared to the one from ID. In this work, we introduce a new model, namely \textit{Bigeminal Priors Variational auto-encoder (BPVAE)}, to address this phenomenon. The BPVAE aims to enhance the robustness of the VAEs by combing the power of VAE with the two independent priors that belong to the training dataset and simple dataset, which complexity is lower than the training dataset, respectively. BPVAE learns two datasets’ features, assigning a higher likelihood for the training dataset than the simple dataset. In this way, we can use BPVAE’s density estimate for detecting the OOD samples. Quantitative experimental results suggest that our model has better generalization capability and stronger robustness than the standard VAEs, proving the effectiveness of the proposed approach of hybrid learning by collaborative priors. Overall, this work paves a new avenue to potentially overcome the OOD problem via multiple latent priors modeling.
Withdrawn
ICLR.cc/2023/Conference
Neural Groundplans: Persistent Neural Scene Representations from a Single Image
We present a method to map 2D image observations of a scene to a persistent 3D scene representation, enabling novel view synthesis and disentangled representation of the movable and immovable components of the scene. Motivated by the bird’s-eye-view (BEV) representation commonly used in vision and robotics, we propose conditional neural groundplans, ground-aligned 2D feature grids, as persistent and memory-efficient scene representations. Our method is trained self-supervised from unlabeled multi-view observations using differentiable rendering, and learns to complete geometry and appearance of occluded regions. In addition, we show that we can leverage multi-view videos at training time to learn to separately reconstruct static and movable components of the scene from a single image at test time. The ability to separately reconstruct movable objects enables a variety of downstream tasks using simple heuristics, such as extraction of object-centric 3D representations, novel view synthesis, instance-level segmentation, 3D bounding box prediction, and scene editing. This highlights the value of neural groundplans as a backbone for efficient 3D scene understanding models.
Accept: poster
ICLR.cc/2018/Conference
Multi-Advisor Reinforcement Learning
We consider tackling a single-agent RL problem by distributing it to $n$ learners. These learners, called advisors, endeavour to solve the problem from a different focus. Their advice, taking the form of action values, is then communicated to an aggregator, which is in control of the system. We show that the local planning method for the advisors is critical and that none of the ones found in the literature is flawless: the \textit{egocentric} planning overestimates values of states where the other advisors disagree, and the \textit{agnostic} planning is inefficient around danger zones. We introduce a novel approach called \textit{empathic} and discuss its theoretical aspects. We empirically examine and validate our theoretical findings on a fruit collection task.
Reject
ICLR.cc/2020/Conference
Bayesian Inference for Large Scale Image Classification
Bayesian inference promises to ground and improve the performance of deep neural networks. It promises to be robust to overfitting, to simplify the training procedure and the space of hyperparameters, and to provide a calibrated measure of uncertainty that can enhance decision making, agent exploration and prediction fairness. Markov Chain Monte Carlo (MCMC) methods enable Bayesian inference by generating samples from the posterior distribution over model parameters. Despite the theoretical advantages of Bayesian inference and the similarity between MCMC and optimization methods, the performance of sampling methods has so far lagged behind optimization methods for large scale deep learning tasks. We aim to fill this gap and introduce ATMC, an adaptive noise MCMC algorithm that estimates and is able to sample from the posterior of a neural network. ATMC dynamically adjusts the amount of momentum and noise applied to each parameter update in order to compensate for the use of stochastic gradients. We use a ResNet architecture without batch normalization to test ATMC on the Cifar10 benchmark and the large scale ImageNet benchmark and show that, despite the absence of batch normalization, ATMC outperforms a strong optimization baseline in terms of both classification accuracy and test log-likelihood. We show that ATMC is intrinsically robust to overfitting on the training data and that ATMC provides a better calibrated measure of uncertainty compared to the optimization baseline.
Reject
ICLR.cc/2021/Conference
Improving Tail Label Prediction for Extreme Multi-label Learning
Extreme multi-label learning (XML) works to annotate objects with relevant labels from an extremely large label set. Many previous methods treat labels uniformly such that the learned model tends to perform better on head labels, while the performance is severely deteriorated for tail labels. However, it is often desirable to predict more tail labels in many real-world applications. To alleviate this problem, in this work, we show theoretical and experimental evidence for the inferior performance of representative XML methods on tail labels. Our finding is that the norm of label classifier weights typically follows a long-tailed distribution similar to the label frequency, which results in the over-suppression of tail labels. Base on this new finding, we present two new modules: (1)~\algoa~learns to re-rank the predictions by optimizing a population-aware loss, which predicts tail labels with high rank; (2)~\algob~augments tail labels via a decoupled learning scheme, which can yield more balanced classification boundary. We conduct experiments on commonly used XML benchmarks with hundreds of thousands of labels, showing that the proposed methods improve the performance of many state-of-the-art XML models by a considerable margin (6\% performance gain with respect to PSP@1 on average).
Reject
ICLR.cc/2022/Conference
Conditional Expectation based Value Decomposition for Scalable On-Demand Ride Pooling
Owing to the benefits for customers (lower prices), drivers (higher revenues), aggregation companies (higher revenues) and the environment (fewer vehicles), on-demand ride pooling (e.g., Uber pool, Grab Share) has become quite popular. The significant computational complexity of matching vehicles to combinations of requests has meant that traditional ride pooling approaches are myopic in that they do not consider the impact of current matches on future value for vehicles/drivers. Recently, Neural Approximate Dynamic Programming (NeurADP) has employed value decomposition with Approximate Dynamic Programming (ADP) to outperform leading approaches by considering the impact of an individual agent's (vehicle) chosen actions on the future value of that agent. However, in order to ensure scalability and facilitate city-scale ride pooling, NeurADP completely ignores the impact of other agents actions on individual agent/vehicle value. As demonstrated in our experimental results, ignoring the impact of other agents actions on individual value can have a significant impact on the overall performance when there is increased competition among vehicles for demand. Our key contribution is a novel mechanism based on computing conditional expectations through joint conditional probabilities for capturing dependencies on other agents actions without increasing the complexity of training or decision making. We show that our new approach, Conditional Expectation based Value Decomposition (CEVD) outperforms NeurADP by up to 9.76$\% $in terms of overall requests served, which is a significant improvement on a city wide benchmark taxi dataset.
Reject
ICLR.cc/2023/Conference
Categorial Grammar Induction as a Compositionality Measure for Emergent Languages in Signaling Games
This paper proposes a method to analyze the compositional structure of emergent languages using Categorial Grammar Induction (CGI). Emergent languages are communication protocols arising among agents in environments such as signaling games. Previous work has studied how similar or dissimilar emergent languages are to natural languages in compositionality. However, most of them focused on trivial compositionality, assuming flat structures in languages. We further focus on non-trivial compositionality, i.e., the relationship between hierarchical syntax and semantics. To this end, we apply CGI to emergent languages, inspired by previous NLP work. Given sentence-meaning pairs of a language, CGI induces 1) a categorial grammar that describes the syntax of the language and 2) a semantic parser that compositionally maps sentences to meanings. We also propose compositionality measures based on the grammar size and semantic parser performance. CGI and the proposed measures enable deeper insights into the non-trivial compositionality of emergent languages, while correlating well with existing measures like TopSim.
Withdrawn
ICLR.cc/2018/Conference
Towards Effective GANs for Data Distributions with Diverse Modes
Generative Adversarial Networks (GANs), when trained on large datasets with diverse modes, are known to produce conflated images which do not distinctly belong to any of the modes. We hypothesize that this problem occurs due to the interaction between two facts: (1) For datasets with large variety, it is likely that the modes lie on separate manifolds. (2) The generator (G) is formulated as a continuous function, and the input noise is derived from a connected set, due to which G's output is a connected set. If G covers all modes, then there must be some portion of G's output which connects them. This corresponds to undesirable, conflated images. We develop theoretical arguments to support these intuitions. We propose a novel method to break the second assumption via learnable discontinuities in the latent noise space. Equivalently, it can be viewed as training several generators, thus creating discontinuities in the G function. We also augment the GAN formulation with a classifier C that predicts which noise partition/generator produced the output images, encouraging diversity between each partition/generator. We experiment on MNIST, celebA, STL-10, and a difficult dataset with clearly distinct modes, and show that the noise partitions correspond to different modes of the data distribution, and produce images of superior quality.
Invite to Workshop Track
ICLR.cc/2020/Conference
FEW-SHOT LEARNING ON GRAPHS VIA SUPER-CLASSES BASED ON GRAPH SPECTRAL MEASURES
We propose to study the problem of few-shot graph classification in graph neural networks (GNNs) to recognize unseen classes, given limited labeled graph examples. Despite several interesting GNN variants being proposed recently for node and graph classification tasks, when faced with scarce labeled examples in the few-shot setting, these GNNs exhibit significant loss in classification performance. Here, we present an approach where a probability measure is assigned to each graph based on the spectrum of the graph’s normalized Laplacian. This enables us to accordingly cluster the graph base-labels associated with each graph into super-classes, where the L^p Wasserstein distance serves as our underlying distance metric. Subsequently, a super-graph constructed based on the super-classes is then fed to our proposed GNN framework which exploits the latent inter-class relationships made explicit by the super-graph to achieve better class label separation among the graphs. We conduct exhaustive empirical evaluations of our proposed method and show that it outperforms both the adaptation of state-of-the-art graph classification methods to few-shot scenario and our naive baseline GNNs. Additionally, we also extend and study the behavior of our method to semi-supervised and active learning scenarios.
Accept (Poster)
ICLR.cc/2023/Conference
CONTINUAL MODEL EVOLVEMENT WITH INNER-PRODUCT RESTRICTION
With pre-trained model's rapid deployment in natural language processing (NLP) applications, it is intuitive to expect these models can continuously evolve when the task requires more complicated inference ability of the model. Existing continual learning (CL) problem setups and methods focus on fixing out-of-distribution (OOD) data streams which cannot solve such a new challenge. We propose a continual model evolvement problem formulation (CME) that introduces a new challenge for fine-tuned pre-trained models that requires them to evolve during deployment. We formulate the problem and introduce multiple metrics to assess current CL methods from different aspects. Further, we propose a strong method dubbed inner-product restriction as a headstart in solving the CME problem. Experimental results indicate that the CME is still challenging to current deployed pre-trained models while our proposed method can provide a strong boost based on previous CL methods, supporting that it is of great need to explore the CME challenge for better deployment of pre-trained models in NLP applications.
Withdrawn
ICLR.cc/2022/Conference
Meta Learning Low Rank Covariance Factors for Energy Based Deterministic Uncertainty
Numerous recent works utilize bi-Lipschitz regularization of neural network layers to preserve relative distances between data instances in the feature spaces of each layer. This distance sensitivity with respect to the data aids in tasks such as uncertainty calibration and out-of-distribution (OOD) detection. In previous works, features extracted with a distance sensitive model are used to construct feature covariance matrices which are used in deterministic uncertainty estimation or OOD detection. However, in cases where there is a distribution over tasks, these methods result in covariances which are sub-optimal, as they may not leverage all of the meta information which can be shared among tasks. With the use of an attentive set encoder, we propose to meta learn either diagonal or diagonal plus low-rank factors to efficiently construct task specific covariance matrices. Additionally, we propose an inference procedure which utilizes scaled energy to achieve a final predictive distribution which is well calibrated under a distributional dataset shift.
Accept (Poster)
ICLR.cc/2023/Conference
Lmser-pix2seq: Learning Stable Sketch Representations For Sketch Healing
Sketch healing aims to recreate a complete sketch from the corrupted one. The sparse and abstract nature of the sketch makes it challenging due to the difficulty in learning. The features extracted from the corrupted sketch may be inconsistent with the ones from the corresponding full sketch. In this paper, we present Lmser-pix2seq to learn stable sketch representations against the missing information by employing a Least mean square error reconstruction (Lmser) block, which falls into encoder-decoder paradigm. Taking as input a corrupted sketch, the Lmser encoder computes the embeddings of structural patterns of the input, while the decoder reconstructs the complete sketch from the embeddings. We build bi-directional skip connections between the encoder and the decoder in our Lmser block. The feedback connections enable recurrent paths to receive more information about the reconstructed sketch produced by the decoder, which helps the encoder extract stable sketch features. The features captured by the Lmser block are eventually fed into a recurrent neural network decoder to recreate the sketches. Experimental results show that our Lmser-pix2seq outperforms the state-of-the-art methods in sketch healing, especially when the sketches are heavily masked or corrupted.
Reject
ICLR.cc/2020/Conference
Scale-Equivariant Steerable Networks
The effectiveness of Convolutional Neural Networks (CNNs) has been substantially attributed to their built-in property of translation equivariance. However, CNNs do not have embedded mechanisms to handle other types of transformations. In this work, we pay attention to scale changes, which regularly appear in various tasks due to the changing distances between the objects and the camera. First, we introduce the general theory for building scale-equivariant convolutional networks with steerable filters. We develop scale-convolution and generalize other common blocks to be scale-equivariant. We demonstrate the computational efficiency and numerical stability of the proposed method. We compare the proposed models to the previously developed methods for scale equivariance and local scale invariance. We demonstrate state-of-the-art results on the MNIST-scale dataset and on the STL-10 dataset in the supervised learning setting.
Accept (Poster)
ICLR.cc/2023/Conference
Self-Consistent Learning: Cooperation between Generators and Discriminators
Using generated data to improve the performance of downstream discriminative models has recently gained popularity due to the great development of pre-trained language models. In most previous studies, generative models and discriminative models are trained separately and thus could not adapt to any changes in each other. As a result, the generated samples can easily deviate from the real data distribution, while the improvement of the discriminative model quickly reaches saturation. Generative adversarial networks (GANs) train generative models via an adversarial process with discriminative models to achieve joint training. However, the training of standard GANs is notoriously unstable and often falls short of convergence. In this paper, to address these issues, we propose a $\textit{self-consistent learning}$ framework, in which a discriminator and a generator are cooperatively trained in a closed-loop form. The discriminator and the generator enhance each other during multiple rounds of alternating training until a scoring consensus is reached. This framework proves to be easy to train and free from instabilities such as mode collapse and non-convergence. Extensive experiments on sentence semantic matching demonstrate the effectiveness of the proposed framework: the discriminator achieves 10+ AP of improvement on the zero-shot setting and new state-of-the-art performance on the full-data setting.
Reject
ICLR.cc/2019/Conference
Mol-CycleGAN - a generative model for molecular optimization
Designing a molecule with desired properties is one of the biggest challenges in drug development, as it requires optimization of chemical compound structures with respect to many complex properties. To augment the compound design process we introduce Mol-CycleGAN -- a CycleGAN-based model that generates optimized compounds with a chemical scaffold of interest. Namely, given a molecule our model generates a structurally similar one with an optimized value of the considered property. We evaluate the performance of the model on selected optimization objectives related to structural properties (presence of halogen groups, number of aromatic rings) and to a physicochemical property (penalized logP). In the task of optimization of penalized logP of drug-like molecules our model significantly outperforms previous results.
Reject
ICLR.cc/2020/Conference
Continual Learning using the SHDL Framework with Skewed Replay Distributions
Human and animals continuously acquire, adapt as well as transfer knowledge throughout their lifespan. The ability to learn continuously is crucial for the effective functioning of agents interacting with the real world and processing continuous streams of information. Continuous learning has been a long-standing challenge for neural networks as the repeated acquisition of information from non-uniform data distributions generally lead to catastrophic forgetting or interference. This work proposes a modular architecture capable of continuous acquisition of tasks while averting catastrophic forgetting. Specifically, our contributions are: (i) Efficient Architecture: a modular architecture emulating the visual cortex that can learn meaningful representations with limited labelled examples, (ii) Knowledge Retention: retention of learned knowledge via limited replay of past experiences, (iii) Forward Transfer: efficient and relatively faster learning on new tasks, and (iv) Naturally Skewed Distributions: The learning in the above-mentioned claims is performed on non-uniform data distributions which better represent the natural statistics of our ongoing experience. Several experiments that substantiate the above-mentioned claims are demonstrated on the CIFAR-100 dataset.
Reject
ICLR.cc/2023/Conference
Git Re-Basin: Merging Models modulo Permutation Symmetries
The success of deep learning is due in large part to our ability to solve certain massive non-convex optimization problems with relative ease. Though non-convex optimization is NP-hard, simple algorithms -- often variants of stochastic gradient descent -- exhibit surprising effectiveness in fitting large neural networks in practice. We argue that neural network loss landscapes often contain (nearly) a single basin after accounting for all possible permutation symmetries of hidden units a la Entezari et al. 2021. We introduce three algorithms to permute the units of one model to bring them into alignment with a reference model in order to merge the two models in weight space. This transformation produces a functionally equivalent set of weights that lie in an approximately convex basin near the reference model. Experimentally, we demonstrate the single basin phenomenon across a variety of model architectures and datasets, including the first (to our knowledge) demonstration of zero-barrier linear mode connectivity between independently trained ResNet models on CIFAR-10. Additionally, we identify intriguing phenomena relating model width and training time to mode connectivity. Finally, we discuss shortcomings of the linear mode connectivity hypothesis, including a counterexample to the single basin theory.
Accept: notable-top-5%
ICLR.cc/2023/Conference
Simplicity bias leads to amplified performance disparities
The simple idea that not all things are equally difficult has surprising implications when applied in a fairness context. In this work we explore how "difficulty" is model-specific, such that different models find different parts of a dataset challenging. When difficulty correlates with group information, we term this difficulty disparity. Drawing a connection with recent work exploring the inductive bias towards simplicity of SGD-trained models, we show that when such a disparity exists, it is further amplified by commonly-used models. We quantify this amplification factor across a range of settings aiming towards a fuller understanding of the role of model bias. We also present a challenge to the simplifying assumption that ``fixing'' a dataset is sufficient to ensure unbiased performance.
Reject
ICLR.cc/2023/Conference
Re-weighting Based Group Fairness Regularization via Classwise Robust Optimization
Many existing group fairness-aware training methods aim to achieve the group fairness by either re-weighting underrepresented groups based on certain rules or using weakly approximated surrogates for the fairness metrics in the objective as regularization terms. Although each of the learning schemes has its own strength in terms of applicability or performance, respectively, it is difficult for any method in the either category to be considered as a gold standard since their successful performances are typically limited to specific cases. To that end, we propose a principled method, dubbed as FairDRO, which unifies the two learning schemes by incorporating a well-justified group fairness metric into the training objective using a classwise distributionally robust optimization (DRO) framework. We then develop an iterative optimization algorithm that minimizes the resulting objective by automatically producing the correct re-weights for each group. Our experiments show that FairDRO is scalable and easily adaptable to diverse applications, and consistently achieves the state-of-the-art performance on several benchmark datasets in terms of the accuracy-fairness trade-off, compared to recent strong baselines.
Accept: poster
ICLR.cc/2022/Conference
A Multi-Task Learning Algorithm for Non-personalized Recommendations
In this paper, we introduce a multi-task learning (MTL) algorithm for recommending non-personalized videos to watch next on industrial video sharing platforms. Personalized recommendations have been studied for decades, while researches on non-personalized solutions are very rare to be seen, which still remain a huge portion in industry. As an indispensable part in recommender system, non-personalized video recommender system also faces several real-world challenges, including maintaining high relevance between source item and target items, as well as achieving multiple competing ranking objectives. To solve these, we largely extended model-based collaborative filtering algorithm by adding related candidate generation stage, Two-tower DNN structure and a multi-task learning mechanism. Compared with typical baseline solutions, our proposed algorithm can capture both linear and non-linear relationships from user-item interactions, and live experiments demonstrate that it can significantly advance the state of the art on recommendation quality.
Desk_Rejected
ICLR.cc/2023/Conference
Pareto Manifold Learning: Tackling multiple tasks via ensembles of single-task models
In Multi-Task Learning, tasks may compete and limit the performance achieved on each other rather than guiding the optimization trajectory to a common solution, superior to its single-task counterparts. There is often not a single solution that is optimal for all tasks, leading practitioners to balance tradeoffs between tasks' performance, and to resort to optimality in the Pareto sense. Current Multi-Task Learning methodologies either completely neglect this aspect of functional diversity, and produce one solution in the Pareto Front predefined by their optimization schemes, or produce diverse but discrete solutions, each requiring a separate training run. In this paper, we conjecture that there exist Pareto Subspaces, i.e., weight subspaces where multiple optimal functional solutions lie. We propose Pareto Manifold Learning, an ensembling method in weight space that is able to discover such a parameterization and produces a continuous Pareto Front in a single training run, allowing practitioners to modulate the performance on each task during inference on the fly. We validate the proposed method on a diverse set of multi-task learning benchmarks, ranging from image classification to tabular datasets and scene understanding, and show that Pareto Manifold Learning outperforms state-of-the-art algorithms.
Reject
ICLR.cc/2023/Conference
Learning Group Importance using the Differentiable Hypergeometric Distribution
Partitioning a set of elements into subsets of a priori unknown sizes is essential in many applications. These subset sizes are rarely explicitly learned - be it the cluster sizes in clustering applications or the number of shared versus independent generative latent factors in weakly-supervised learning. Probability distributions over correct combinations of subset sizes are non-differentiable due to hard constraints, which prohibit gradient-based optimization. In this work, we propose the differentiable hypergeometric distribution. The hypergeometric distribution models the probability of different group sizes based on their relative importance. We introduce reparameterizable gradients to learn the importance between groups and highlight the advantage of explicitly learning the size of subsets in two typical applications: weakly-supervised learning and clustering. In both applications, we outperform previous approaches, which rely on suboptimal heuristics to model the unknown size of groups.
Accept: notable-top-25%
ICLR.cc/2022/Conference
Graph-Enhanced Exploration for Goal-oriented Reinforcement Learning
Goal-oriented Reinforcement Learning (GoRL) is a promising approach for scaling up RL techniques on sparse reward environments requiring long horizon planning. Recent works attempt to build suitable abstraction graph of the environment and enhance GoRL with classical graphical methods such as shortest path searching; however, these approaches mainly focus on either graph construction or agent exploitation, but leave the exploration lack of study. This paper proposes Graph-enhanced GoRL (G2RL), a new GoRL framework for effective exploration and efficient training based on the state-transition graph. We first introduce the optimal goals for exploration on the graph and then use them as supervised signals to train the goal generator in G2RL in a hindsight manner. Furthermore, we define relevant trajectories of a state based on its graph neighborhood and show that giving high priority to these trajectories would lead to an efficient policy learning. In addition to the theoretical results regarding optimal goal generation, our empirical results on standard discrete and continuous control benchmarks show that leveraging the state-transition graph is beneficial for GoRL to learn an effective and informative exploration strategy and outperform the state-of-the-art methods.
Withdrawn
ICLR.cc/2020/Conference
Kernel of CycleGAN as a principal homogeneous space
Unpaired image-to-image translation has attracted significant interest due to the invention of CycleGAN, a method which utilizes a combination of adversarial and cycle consistency losses to avoid the need for paired data. It is known that the CycleGAN problem might admit multiple solutions, and our goal in this paper is to analyze the space of exact solutions and to give perturbation bounds for approximate solutions. We show theoretically that the exact solution space is invariant with respect to automorphisms of the underlying probability spaces, and, furthermore, that the group of automorphisms acts freely and transitively on the space of exact solutions. We examine the case of zero pure CycleGAN loss first in its generality, and, subsequently, expand our analysis to approximate solutions for extended CycleGAN loss where identity loss term is included. In order to demonstrate that these results are applicable, we show that under mild conditions nontrivial smooth automorphisms exist. Furthermore, we provide empirical evidence that neural networks can learn these automorphisms with unexpected and unwanted results. We conclude that finding optimal solutions to the CycleGAN loss does not necessarily lead to the envisioned result in image-to-image translation tasks and that underlying hidden symmetries can render the result useless.
Accept (Poster)
ICLR.cc/2021/Conference
Representation Balancing Offline Model-based Reinforcement Learning
One of the main challenges in offline and off-policy reinforcement learning is to cope with the distribution shift that arises from the mismatch between the target policy and the data collection policy. In this paper, we focus on a model-based approach, particularly on learning the representation for a robust model of the environment under the distribution shift, which has been first studied by Representation Balancing MDP (RepBM). Although this prior work has shown promising results, there are a number of shortcomings that still hinder its applicability to practical tasks. In particular, we address the curse of horizon exhibited by RepBM, rejecting most of the pre-collected data in long-term tasks. We present a new objective for model learning motivated by recent advances in the estimation of stationary distribution corrections. This effectively overcomes the aforementioned limitation of RepBM, as well as naturally extending to continuous action spaces and stochastic policies. We also present an offline model-based policy optimization using this new objective, yielding the state-of-the-art performance in a representative set of benchmark offline RL tasks.
Accept (Poster)
ICLR.cc/2022/Conference
Bootstrapped Hindsight Experience replay with Counterintuitive Prioritization
Goal-conditioned environments are known as sparse rewards tasks, in which the agent gains a positive reward only when it achieves the goal. Such an setting results in much difficulty for the agent to explore successful trajectories. Hindsight experience replay (HER) replaces the goal in failed experiences with any practically achieved one, so that the agent has a much higher chance to see successful trajectories even if they are fake. Comprehensive results have demonstrated the effectiveness of HER in the literature. However, the importance of the fake trajectories differs in terms of exploration and exploitation, and it is usually inefficient to learn with a fixed proportion of fake and original data as HER did. In this paper, inspired by Bootstrapped DQN, we use multiple heads in DDPG and take advantage of the diversity and uncertainty among multiple heads to improve the data efficiency with relabeled goals. The method is referred to as Bootstrapped HER (BHER). Specifically, in addition to the benefit from the Bootstrapped version, we explicitly leverage the uncertainty measured by the variance of estimated Q-values from multiple heads. A common knowledge is that higher uncertainty will promote exploration and hence maximizing the uncertainty via a bonus term will induce better performance in Q-learning. However, in this paper, we reveal a counterintuitive conclusion that for hindsight experiences, exploiting lower uncertainty data samples will significantly improve the performance. The explanation behind this fact is that hindsight relabeling largely promotes exploration, and then exploiting lower uncertainty data (whose goals are generated by hindsight relabeling) provides a good trade-off between exploration and exploitation, resulting in further improved data efficiency. Comprehensive experiments demonstrate that our method can achieve state-of-the-art results in many goal-conditioned tasks.
Reject
ICLR.cc/2020/Conference
Scheduled Intrinsic Drive: A Hierarchical Take on Intrinsically Motivated Exploration
Exploration in sparse reward reinforcement learning remains an open challenge. Many state-of-the-art methods use intrinsic motivation to complement the sparse extrinsic reward signal, giving the agent more opportunities to receive feedback during exploration. Commonly these signals are added as bonus rewards, which results in a mixture policy that neither conducts exploration nor task fulfillment resolutely. In this paper, we instead learn separate intrinsic and extrinsic task policies and schedule between these different drives to accelerate exploration and stabilize learning. Moreover, we introduce a new type of intrinsic reward denoted as successor feature control (SFC), which is general and not task-specific. It takes into account statistics over complete trajectories and thus differs from previous methods that only use local information to evaluate intrinsic motivation. We evaluate our proposed scheduled intrinsic drive (SID) agent using three different environments with pure visual inputs: VizDoom, DeepMind Lab and DeepMind Control Suite. The results show a substantially improved exploration efficiency with SFC and the hierarchical usage of the intrinsic drives. A video of our experimental results can be found at https://gofile.io/?c=HpEwTd.
Reject
ICLR.cc/2022/Conference
Multi-Task Processes
Neural Processes (NPs) consider a task as a function realized from a stochastic process and flexibly adapt to unseen tasks through inference on functions. However, naive NPs can model data from only a single stochastic process and are designed to infer each task independently. Since many real-world data represent a set of correlated tasks from multiple sources (e.g., multiple attributes and multi-sensor data), it is beneficial to infer them jointly and exploit the underlying correlation to improve the predictive performance. To this end, we propose Multi-Task Neural Processes (MTNPs), an extension of NPs designed to jointly infer tasks realized from multiple stochastic processes. We build MTNPs in a hierarchical way such that inter-task correlation is considered by conditioning all per-task latent variables on a single global latent variable. In addition, we further design our MTNPs so that they can address multi-task settings with incomplete data (i.e., not all tasks share the same set of input points), which has high practical demands in various applications. Experiments demonstrate that MTNPs can successfully model multiple tasks jointly by discovering and exploiting their correlations in various real-world data such as time series of weather attributes and pixel-aligned visual modalities. We release our code at https://github.com/GitGyun/multi_task_neural_processes.
Accept (Poster)
ICLR.cc/2023/Conference
CLIP-FLOW: CONTRASTIVE LEARNING WITH ITERATIVE PSEUDO LABELING FOR OPTICAL FLOW
Synthetic datasets are often used to pretrain end-to-end optical flow networks, due to the lack of a large amount of labeled, real scene data. But major drops in accuracy occur when moving from synthetic to real scenes. How do we better transfer the knowledge learned from synthetic to real domains? To this end, we propose CLIP-Flow, a semi-supervised iterative pseudo labeling framework to transfer the pretraining knowledge to the target real domain. We leverage large-scale, unlabeled real data to facilitate transfer learning with the supervision of iteratively updated pseudo ground truth labels, bridging the domain gap between the synthetic and the real. In addition, we propose a contrastive flow loss on reference features and the warped features by pseudo ground truth flows, to further boost the accurate matching and dampen the mismatching due to motion, occlusion, or noisy pseudo labels. We adopt RAFT as the backbone and obtain an F1-all error of 4.11%, i.e., a 19% error reduction from RAFT (5.10%) and ranking 2nd place at submission on KITTI 2015 benchmark. Our framework can also be extended to other models, e.g., CRAFT, reducing the F1-all error from 4.79% to 4.66% on KITTI 2015 benchmark.
Reject
ICLR.cc/2021/Conference
Optimizing Transformers with Approximate Computing for Faster, Smaller and more Accurate NLP Models
Transformer models have garnered a lot of interest in recent years by delivering state-of-the-art performance in a range of Natural Language Processing (NLP) tasks. However, these models can have over a hundred billion parameters, presenting very high computational and memory requirements. We address this challenge through Approximate Computing, specifically targeting the use of Transformers in NLP tasks. Transformers are typically pre-trained and subsequently specialized for specific tasks through transfer learning. We observe that pre-trained Transformers are often over-parameterized for several downstream NLP tasks and propose a framework to create smaller and faster models with comparable accuracy. The key cornerstones of the framework are a Significance Analysis (SA) method to identify important components in a pre-trained Transformer for a given task, and techniques to approximate the less significant components. Our framework can be adapted to produce models that are faster, smaller and/or more accurate, depending on the user's constraints. We apply our framework to multiple Transformer models and different downstream tasks, including previously proposed optimized models like DistilBERT and Q8BERT. We demonstrate that our framework produces models that are up to 4$\times$ faster and up to 14$\times$ smaller (with less than 0.5% relative accuracy degradation), or up to 5.5% more accurate with simultaneous model size and speed improvements of up to 9.8$\times$ and 2.9$\times$, respectively.
Reject
ICLR.cc/2021/Conference
Understanding Mental Representations Of Objects Through Verbs Applied To Them
In order to interact with objects in our environment, we rely on an understanding of the actions that can be performed on them, and the extent to which they rely or have an effect on the properties of the object. This knowledge is called the object "affordance". We propose an approach for creating an embedding of objects in an affordance space, in which each dimension corresponds to an aspect of meaning shared by many actions, using text corpora. This embedding makes it possible to predict which verbs will be applicable to a given object, as captured in human judgments of affordance, better than a variety of alternative approaches. Furthermore, we show that the dimensions learned are interpretable, and that they correspond to typical patterns of interaction with objects. Finally, we show that the dimensions can be used to predict a state-of-the-art mental representation of objects, derived purely from human judgements of object similarity.
Reject
ICLR.cc/2023/Conference
Improving Vision Attention with Random Walk Graph Kernel
Vision transformers, which propose to tokenize an image and introduce attention mechanism to learn cross-token relationship, have advanced many computer vision tasks.However, the attention module owns a quadratic computational complexity and hence suffers from slow computing speed and high memory cost, hindering it from handling long sequences of tokens.Some attempts optimize the quadratic attention with linear approximation yet observe undesired performance drop.This work balances the trade-off between modeling efficiency and capacity of vision attention.We notice that, by treating queries and keys as nodes in a graph, existing algorithms are akin to modeling one-step interaction between nodes.To strengthen the cross-node connection for a more representative attention, we introduce multi-step interaction, which is equivalent to solving an inverse matrix as in random walk graph kernel.We then come up with a new strategy to construct queries and keys, with the help of bipartite graph, to ease the calculation of matrix inversion.The effectiveness of our approach is verified on various visual tasks. We also make it possible to learn a vision transformer with extremely long sequences of tokens.We achieved the competitive results on the semantic segmentation task with 15% fewer parameters and 10-25% less computation. In addition, the vision transformer based quantization method can be applied to 512x512 or even 1024x1024 resolution images. Code will be made publicly available.
Withdrawn
ICLR.cc/2023/Conference
Global View For GCN: Why Go Deep When You Can Be Shallow?
Existing graph convolutional network (GCN) methods attempt to expand the receptive field of its convolution by either stacking up more convolutional layers or accumulating multi-hop adjacency matrices. Either approach increases computation complexity while providing a limited view of the network topology. We propose to extend k-hop adjacency matrices into one generalized exponential matrix to provide GCNs with a global overview of the network topology. This technique allows the GCNs to learn global topology without going deep and with much fewer parameters than most state-of-the-art GCNs, challenging the common assumption that deep GCNs are empirically better for learning global features. We show a significant improvement in performance in semi-supervised learning when this technique is used for common GCNs while maintaining much shallower network architectures ($\leq4$ layers) than the existing ones.
Withdrawn
ICLR.cc/2023/Conference
Sampling with Mollified Interaction Energy Descent
Sampling from a target measure whose density is only known up to a normalization constant is a fundamental problem in computational statistics and machine learning. In this paper, we present a new optimization-based method for sampling called mollified interaction energy descent (MIED). MIED minimizes a new class of energies on probability measures called mollified interaction energies (MIEs). These energies rely on mollifier functions---smooth approximations of the Dirac delta originated from PDE theory. We show that as the mollifier approaches the Dirac delta, the MIE converges to the chi-square divergence with respect to the target measure and the gradient flow of the MIE agrees with that of the chi-square divergence. Optimizing this energy with proper discretization yields a practical first-order particle-based algorithm for sampling in both unconstrained and constrained domains. We show experimentally that for unconstrained sampling problems our algorithm performs on par with existing particle-based algorithms like SVGD, while for constrained sampling problems our method readily incorporates constrained optimization techniques to handle more flexible constraints with strong performance compared to alternatives.
Accept: poster
ICLR.cc/2021/Conference
Uncertainty Sets for Image Classifiers using Conformal Prediction
Convolutional image classifiers can achieve high predictive accuracy, but quantifying their uncertainty remains an unresolved challenge, hindering their deployment in consequential settings. Existing uncertainty quantification techniques, such as Platt scaling, attempt to calibrate the network’s probability estimates, but they do not have formal guarantees. We present an algorithm that modifies any classifier to output a predictive set containing the true label with a user-specified probability, such as 90%. The algorithm is simple and fast like Platt scaling, but provides a formal finite-sample coverage guarantee for every model and dataset. Our method modifies an existing conformal prediction algorithm to give more stable predictive sets by regularizing the small scores of unlikely classes after Platt scaling. In experiments on both Imagenet and Imagenet-V2 with ResNet-152 and other classifiers, our scheme outperforms existing approaches, achieving coverage with sets that are often factors of 5 to 10 smaller than a stand-alone Platt scaling baseline.
Accept (Spotlight)
ICLR.cc/2022/Conference
Causal Contextual Bandits with Targeted Interventions
We study a contextual bandit setting where the learning agent has the ability to perform interventions on targeted subsets of the population, apart from possessing qualitative causal side-information. This novel formalism captures intricacies in real-world scenarios such as software product experimentation where targeted experiments can be conducted. However, this fundamentally changes the set of options that the agent has, compared to standard contextual bandit settings, necessitating new techniques. This is also the first work that integrates causal side-information in a contextual bandit setting, where the agent aims to learn a policy that maps contexts to arms (as opposed to just identifying one best arm). We propose a new algorithm, which we show empirically performs better than baselines on experiments that use purely synthetic data and on real world-inspired experiments. We also prove a bound on regret that theoretically guards performance.
Accept (Poster)
ICLR.cc/2023/Conference
Confounder Identification-free Causal Visual Feature Learning
Confounders in deep learning are in general detrimental to model's generalization where they infiltrate feature representations. Therefore, learning causal features that are free of interference from confounders is important. Most previous causal learning based approaches employ back-door criterion to mitigate the adverse effect of certain specific confounder, which require the explicit identification of confounder. However, in real scenarios, confounders are typically diverse and difficult to be identified. In this paper, we propose a novel Confounder Identification-free Causal Visual Feature Learning (CICF) method, which obviates the need for identifying confounders. CICF models the interventions among different samples based on front-door criterion, and then approximates the global-scope intervening effect upon the instance-level interventions from the perspective of optimization. In this way, we aim to find a reliable optimization direction, which avoids the intervening effects of confounders, to learn causal features. Furthermore, we uncover the relation between CICF and the popular meta-learning strategy MAML, and provide an interpretation of why MAML works from the theoretical perspective of causal learning for the first time. Thanks to the effective learning of causal features, our CICF enables models to have superior generalization capability. Extensive experiments on domain generalization benchmark datasets demonstrate the effectiveness of our CICF, which achieves the state-of-the-art performance.
Reject
ICLR.cc/2023/Conference
Lovasz Theta Contrastive Learning
We establish a connection between the Lovasz theta function of a graph and the widely used InfoNCE loss. We show that under certain conditions, the minima of the InfoNCE loss are related to minimizing the Lovasz theta function on the empty similarity graph between the samples. Building on this connection, we generalize contrastive learning on weighted similarity graphs between samples. Our Lovasz theta contrastive loss uses a weighted graph that can be learned to take into account similarities between our data. We evaluate our method on image classification tasks, demonstrating an improvement of $1 \%$ in the supervised case and up to $4 \%$ in the unsupervised case.
Reject
ICLR.cc/2018/Conference
Variational Inference of Disentangled Latent Concepts from Unlabeled Observations
Disentangled representations, where the higher level data generative factors are reflected in disjoint latent dimensions, offer several benefits such as ease of deriving invariant representations, transferability to other tasks, interpretability, etc. We consider the problem of unsupervised learning of disentangled representations from large pool of unlabeled observations, and propose a variational inference based approach to infer disentangled latent factors. We introduce a regularizer on the expectation of the approximate posterior over observed data that encourages the disentanglement. We also propose a new disentanglement metric which is better aligned with the qualitative disentanglement observed in the decoder's output. We empirically observe significant improvement over existing methods in terms of both disentanglement and data likelihood (reconstruction quality).
Accept (Poster)
ICLR.cc/2021/Conference
R-GAP: Recursive Gradient Attack on Privacy
Federated learning frameworks have been regarded as a promising approach to break the dilemma between demands on privacy and the promise of learning from large collections of distributed data. Many such frameworks only ask collaborators to share their local update of a common model, i.e. gradients with respect to locally stored data, instead of exposing their raw data to other collaborators. However, recent optimization-based gradient attacks show that raw data can often be accurately recovered from gradients. It has been shown that minimizing the Euclidean distance between true gradients and those calculated from estimated data is often effective in fully recovering private data. However, there is a fundamental lack of theoretical understanding of how and when gradients can lead to unique recovery of original data. Our research fills this gap by providing a closed-form recursive procedure to recover data from gradients in deep neural networks. We name it Recursive Gradient Attack on Privacy (R-GAP). Experimental results demonstrate that R-GAP works as well as or even better than optimization-based approaches at a fraction of the computation under certain conditions. Additionally, we propose a Rank Analysis method, which can be used to estimate the risk of gradient attacks inherent in certain network architectures, regardless of whether an optimization-based or closed-form-recursive attack is used. Experimental results demonstrate the utility of the rank analysis towards improving the network's security. Source code is available for download from https://github.com/JunyiZhu-AI/R-GAP.
Accept (Poster)
ICLR.cc/2022/Conference
Diverse and Consistent Multi-view Networks for Semi-supervised Regression
Label collection is costly in many applications, which poses the need for label-efficient learning. In this work, we present Diverse and Consistent Multi-view Networks (DiCoM) — a novel semi-supervised regression technique based on a multi-view learning framework. DiCoM combines diversity with consistency — two seemingly opposing yet complementary principles of multi-view learning - based on underlying probabilistic graphical assumptions. Given multiple deep views of the same input, DiCoM encourages a negative correlation among the views' predictions on labeled data, while simultaneously enforces their agreement on unlabeled data. DiCoM can utilize either multi-network or multi-branch architectures to make a trade-off between computational cost and modeling performance. Under realistic evaluation setups, DiCoM outperforms competing methods on tabular and image data. Our ablation studies confirm the importance of having both consistency and diversity.
Reject
ICLR.cc/2023/Conference
Restoration based Generative Models
Denoising generative models (DGMs) have recently attracted increasing attention by showing impressive synthesis quality. DGMs are built on a diffusion process that pushes data to the noise distribution and the models learn to denoise. In this paper, we establish the interpretation of DGMs in terms of image restoration (IR). Integrating IR literature allows us to use an alternative objective and diverse forward processes, not confining to the diffusion process. By imposing prior knowledge on the loss function grounded on MAP estimation, we eliminate the need for the expensive sampling of DGMs. Also, we propose a multi-scale training, which alleviates the latent inefficiency of DGMs, by taking advantage of the flexibility of the forward process. Our model improves the quality and efficiency of both training and inference, achieving state-of-the-art performance when the number of forward steps is limited. Furthermore, we show the applicability of our model to inverse problems. We believe that our framework paves the way for designing a new type of flexible general generative model.
Reject
ICLR.cc/2021/Conference
Decoupling Representation Learning from Reinforcement Learning
In an effort to overcome limitations of reward-driven feature learning in deep reinforcement learning (RL) from images, we propose decoupling representation learning from policy learning. To this end, we introduce a new unsupervised learning (UL) task, called Augmented Temporal Contrast (ATC), which trains a convolutional encoder to associate pairs of observations separated by a short time difference, under image augmentations and using a contrastive loss. In online RL experiments, we show that training the encoder exclusively using ATC matches or outperforms end-to-end RL in most environments. Additionally, we benchmark several leading UL algorithms by pre-training encoders on expert demonstrations and using them, with weights frozen, in RL agents; we find that agents using ATC-trained encoders outperform all others. We also train multi-task encoders on data from multiple environments and show generalization to different downstream RL tasks. Finally, we ablate components of ATC, and introduce a new data augmentation to enable replay of (compressed) latent images from pre-trained encoders when RL requires augmentation. Our experiments span visually diverse RL benchmarks in DeepMind Control, DeepMind Lab, and Atari, and our complete code is available at \url{hidden url}.
Reject
ICLR.cc/2019/Conference
Regularized Learning for Domain Adaptation under Label Shifts
We propose Regularized Learning under Label shifts (RLLS), a principled and a practical domain-adaptation algorithm to correct for shifts in the label distribution between a source and a target domain. We first estimate importance weights using labeled source data and unlabeled target data, and then train a classifier on the weighted source samples. We derive a generalization bound for the classifier on the target domain which is independent of the (ambient) data dimensions, and instead only depends on the complexity of the function class. To the best of our knowledge, this is the first generalization bound for the label-shift problem where the labels in the target domain are not available. Based on this bound, we propose a regularized estimator for the small-sample regime which accounts for the uncertainty in the estimated weights. Experiments on the CIFAR-10 and MNIST datasets show that RLLS improves classification accuracy, especially in the low sample and large-shift regimes, compared to previous methods.
Accept (Poster)
ICLR.cc/2023/Conference
Understanding Embodied Reference with Touch-Line Transformer
We study embodied reference understanding, the task of locating referents using embodied gestural signals and language references. Human studies have revealed that, contrary to popular belief, objects referred to or pointed to do not lie on the elbow-wrist line, but rather on the so-called virtual touch line. Nevertheless, contemporary human pose representations lack the virtual touch line. To tackle this problem, we devise the touch-line Transformer: It takes as input tokenized visual and textual features and simultaneously predicts the referent’s bounding box and a touch-line vector. Leveraging this touch-line prior, we further devise a geometric consistency loss that promotes co-linearity between referents and touch lines. Using the touch line as gestural information dramatically improves model performances: Experiments on the YouRefIt dataset demonstrate that our method yields a +25.0% accuracy improvement under the 0.75 IoU criterion, hence closing 63.6% of the performance difference between models and humans. Furthermore, we computationally validate prior human studies by demonstrating that computational models more accurately locate referents when employing the virtual touch line than when using the elbow-wrist line.
Accept: poster
ICLR.cc/2018/Conference
Sensor Transformation Attention Networks
Recent work on encoder-decoder models for sequence-to-sequence mapping has shown that integrating both temporal and spatial attentional mechanisms into neural networks increases the performance of the system substantially. We report on a new modular network architecture that applies an attentional mechanism not on temporal and spatial regions of the input, but on sensor selection for multi-sensor setups. This network called the sensor transformation attention network (STAN) is evaluated in scenarios which include the presence of natural noise or synthetic dynamic noise. We demonstrate how the attentional signal responds dynamically to changing noise levels and sensor-specific noise, leading to reduced word error rates (WERs) on both audio and visual tasks using TIDIGITS and GRID; and also on CHiME-3, a multi-microphone real-world noisy dataset. The improvement grows as more channels are corrupted as demonstrated on the CHiME-3 dataset. Moreover, the proposed STAN architecture naturally introduces a number of advantages including ease of removing sensors from existing architectures, attentional interpretability, and increased robustness to a variety of noise environments.
Reject
ICLR.cc/2019/Conference
Few-Shot Intent Inference via Meta-Inverse Reinforcement Learning
A significant challenge for the practical application of reinforcement learning toreal world problems is the need to specify an oracle reward function that correctly defines a task. Inverse reinforcement learning (IRL) seeks to avoid this challenge by instead inferring a reward function from expert behavior. While appealing, it can be impractically expensive to collect datasets of demonstrations that cover the variation common in the real world (e.g. opening any type of door). Thus in practice, IRL must commonly be performed with only a limited set of demonstrations where it can be exceedingly difficult to unambiguously recover a reward function. In this work, we exploit the insight that demonstrations from other tasks can be used to constrain the set of possible reward functions by learning a "prior" that is specifically optimized for the ability to infer expressive reward functions from limited numbers of demonstrations. We demonstrate that our method can efficiently recover rewards from images for novel tasks and provide intuition as to how our approach is analogous to learning a prior.
Reject
ICLR.cc/2020/Conference
All SMILES Variational Autoencoder for Molecular Property Prediction and Optimization
Variational autoencoders (VAEs) defined over SMILES string and graph-based representations of molecules promise to improve the optimization of molecular properties, thereby revolutionizing the pharmaceuticals and materials industries. However, these VAEs are hindered by the non-unique nature of SMILES strings and the computational cost of graph convolutions. To efficiently pass messages along all paths through the molecular graph, we encode multiple SMILES strings of a single molecule using a set of stacked recurrent neural networks, harmonizing hidden representations of each atom between SMILES representations, and use attentional pooling to build a final fixed-length latent representation. By then decoding to a disjoint set of SMILES strings of the molecule, our All SMILES VAE learns an almost bijective mapping between molecules and latent representations near the high-probability-mass subspace of the prior. Our SMILES-derived but molecule-based latent representations significantly surpass the state-of-the-art in a variety of fully- and semi-supervised property regression and molecular property optimization tasks.
Reject
ICLR.cc/2023/Conference
Diffusion-based Image Translation using disentangled style and content representation
Diffusion-based image translation guided by semantic texts or a single target image has enabled flexible style transfer which is not limited to the specific domains. Unfortunately, due to the stochastic nature of diffusion models, it is often difficult to maintain the original content of the image during the reverse diffusion. To address this, here we present a novel diffusion-based unsupervised image translation method, dubbed as DiffuseIT, using disentangled style and content representation. Specifically, inspired by the slicing Vision Transformer, we extract intermediate keys of multihead self attention layer from ViT model and used them as the content preservation loss. Then, an image guided style transfer is performed by matching the [CLS] classification token from the denoised samples and target image, whereas additional CLIP loss is used for the text-driven style transfer. To further accelerate the semantic change during the reverse diffusion, we also propose a novel semantic divergence loss and resampling strategy. Our experimental results show that the proposed method outperforms state-of-the-art baseline models in both text-guided and image-guided translation tasks.
Accept: poster
ICLR.cc/2023/Conference
Learn the Time to Learn: Replay Scheduling in Continual Learning
Replay methods are known to be successful at mitigating catastrophic forgetting in continual learning scenarios despite having limited access to historical data. However, storing historical data is cheap in many real-world settings, yet replaying all historical data is often prohibited due to processing time constraints. In such settings, we propose that continual learning systems should learn the time to learn and schedule which tasks to replay at different time steps. We first demonstrate the benefits of our proposal by using Monte Carlo tree search to find a proper replay schedule, and show that the found replay schedules can outperform fixed scheduling policies when combined with multiple replay methods in various continual learning settings. Additionally, we propose a framework for learning replay scheduling policies with reinforcement learning. We show that the learned policies can generalize better in new continual learning scenarios compared to equally replaying all seen tasks, without added computational cost. Our study reveals the importance of learning the time to learn in continual learning, which brings current research closer to real-world needs.
Withdrawn
ICLR.cc/2022/Conference
Neural Deep Equilibrium Solvers
A deep equilibrium (DEQ) model abandons traditional depth by solving for the fixed point of a single nonlinear layer $f_\theta$. This structure enables decoupling the internal structure of the layer (which controls representational capacity) from how the fixed point is actually computed (which impacts inference-time efficiency), which is usually via classic techniques such as Broyden's method or Anderson acceleration. In this paper, we show that one can exploit such decoupling and substantially enhance this fixed point computation using a custom neural solver. Specifically, our solver uses a parameterized network to both guess an initial value of the optimization and perform iterative updates, in a method that generalizes a learnable form of Anderson acceleration and can be trained end-to-end in an unsupervised manner. Such a solution is particularly well suited to the implicit model setting, because inference in these models requires repeatedly solving for a fixed point of the same nonlinear layer for different inputs, a task at which our network excels. Our experiments show that these neural equilibrium solvers are fast to train (only taking an extra 0.9-1.1% over the original DEQ's training time), require few additional parameters (1-3% of the original model size), yet lead to a $2\times$ speedup in DEQ network inference without any degradation in accuracy across numerous domains and tasks.
Accept (Poster)
ICLR.cc/2023/Conference
MPCFORMER: FAST, PERFORMANT AND PRIVATE TRANSFORMER INFERENCE WITH MPC
Enabling private inference is crucial for many cloud inference services that are based on Transformer models. However, existing private inference solutions can increase the inference latency by more than 60$\times$ or significantly compromise the inference quality. In this paper, we design the framework MPCFORMER as a practical solution, using Secure Multi-Party Computation (MPC) and Knowledge Distillation (KD). Through extensive evaluations, we show that MPCFORMER significantly speeds up Transformer inference in MPC settings while achieving similar ML performance to the input model. On the IMDb dataset, it achieves similar performance to $\text{BERT}_\text{BASE}$, while being 5.3$\times$ faster. On the GLUE benchmark, it achieves 97% performance of $\text{BERT}_\text{BASE}$ with a 2.2$\times$ speedup. MPCFORMER remains effective with different trained Transformer weights such as $\text{ROBERTA}_\text{BASE}$ and larger models including $\text{BERT}_\text{LARGE}$. Code is available at https://github.com/MccRee177/MPCFormer.
Accept: notable-top-25%
ICLR.cc/2023/Conference
GeONet: a neural operator for learning the Wasserstein geodesic
Optimal transport (OT) offers a versatile framework to compare complex data distributions in a geometrically meaningful way. Traditional methods for computing the Wasserstein distance and geodesic between probability measures require mesh-dependent domain discretization and suffer from the curse-of-dimensionality. We present GeONet, a mesh-invariant deep neural operator network that learns the non-linear mapping from the input pair of initial and terminal distributions to the Wasserstein geodesic connecting the two endpoint distributions. In the offline training stage, GeONet learns the saddle point optimality conditions for the dynamic formulation of the OT problem in the primal and dual spaces that are characterized by a coupled PDE system. The subsequent inference stage is instantaneous and can be deployed for real-time predictions in the online learning setting. We demonstrate that GeONet achieves comparable testing accuracy to the standard OT solvers on a simulation example and the CIFAR-10 dataset with considerably reduced inference-stage computational cost by orders of magnitude.
Withdrawn
ICLR.cc/2022/Conference
How to Train Your MAML to Excel in Few-Shot Classification
Model-agnostic meta-learning (MAML) is arguably one of the most popular meta-learning algorithms nowadays. Nevertheless, its performance on few-shot classification is far behind many recent algorithms dedicated to the problem. In this paper, we point out several key facets of how to train MAML to excel in few-shot classification. First, we find that MAML needs a large number of gradient steps in its inner loop update, which contradicts its common usage in few-shot classification. Second, we find that MAML is sensitive to the class label assignments during meta-testing. Concretely, MAML meta-trains the initialization of an $N$-way classifier. These $N$ ways, during meta-testing, then have "$N!$" different permutations to be paired with a few-shot task of $N$ novel classes. We find that these permutations lead to a huge variance of accuracy, making MAML unstable in few-shot classification. Third, we investigate several approaches to make MAML permutation-invariant, among which meta-training a single vector to initialize all the $N$ weight vectors in the classification head performs the best. On benchmark datasets like MiniImageNet and TieredImageNet, our approach, which we name UNICORN-MAML, performs on a par with or even outperforms many recent few-shot classification algorithms, without sacrificing MAML's simplicity.
Accept (Poster)
ICLR.cc/2021/Conference
Training By Vanilla SGD with Larger Learning Rates
The stochastic gradient descent (SGD) method, first proposed in 1950's, has been the foundation for deep-neural-network (DNN) training with numerous enhancements including adding a momentum or adaptively selecting learning rates, or using both strategies and more. A common view for SGD is that the learning rate should be eventually made small in order to reach sufficiently good approximate solutions. Another widely held view is that the vanilla SGD is out of fashion in comparison to many of its modern variations. In this work, we provide a contrarian claim that, when training over-parameterized DNNs, the vanilla SGD can still compete well with, and oftentimes outperform, its more recent variations by simply using learning rates significantly larger than commonly used values. We establish theoretical results to explain this local convergence behavior of SGD on nonconvex functions, and also present computational evidence, across multiple tasks including image classification, speech recognition and natural language processing, to support the practice of using larger learning rates.
Reject
ICLR.cc/2020/Conference
One Generation Knowledge Distillation by Utilizing Peer Samples
Knowledge Distillation (KD) is a widely used technique in recent deep learning research to obtain small and simple models whose performance is on a par with their large and complex counterparts. Standard Knowledge Distillation tends to be time-consuming because of the training time spent to obtain a teacher model that would then provide guidance for the student model. It might be possible to cut short the time by training a teacher model on the fly, but it is not trivial to have such a high-capacity teacher that gives quality guidance to student models this way. To improve this, we present a novel framework of Knowledge Distillation exploiting dark knowledge from the whole training set. In this framework, we propose a simple and effective implementation named Distillation by Utilizing Peer Samples (DUPS) in one generation. We verify our algorithm on numerous experiments. Compared with standard training on modern architectures, DUPS achieves an average improvement of 1%-2% on various tasks with nearly zero extra cost. Considering some typical Knowledge Distillation methods which are much more time-consuming, we also get comparable or even better performance using DUPS.
Withdrawn
ICLR.cc/2021/Conference
Decomposing Mutual Information for Representation Learning
Many self-supervised representation learning methods maximize mutual information (MI) across views. In this paper, we transform each view into a set of subviews and then decompose the original MI bound into a sum of bounds involving conditional MI between the subviews. E.g.,~given two views $x$ and $y$ of the same input example, we can split $x$ into two subviews, $x^{\prime}$ and $x^{\prime\prime}$, which depend only on $x$ but are otherwise unconstrained. The following holds: $I(x; y) \geq I(x^{\prime\prime}; y) + I(x^{\prime}; y | x^{\prime\prime})$, due to the chain rule and information processing inequality. By maximizing both terms in the decomposition, our approach explicitly rewards the encoder for any information about $y$ which it extracts from $x^{\prime\prime}$, and for information about $y$ extracted from $x^{\prime}$ in excess of the information from $x^{\prime\prime}$. We provide a novel contrastive lower-bound on conditional MI, that relies on sampling contrast sets from $p(y|x^{\prime\prime})$. By decomposing the original MI into a sum of increasingly challenging MI bounds between sets of increasingly informed views, our representations can capture more of the total information shared between the original views. We empirically test the method in a vision domain and for dialogue generation.
Reject
ICLR.cc/2023/Conference
Robust and accelerated single-spike spiking neural network training with applicability to challenging temporal tasks
Spiking neural networks (SNNs), particularly the single-spike variant in which neurons spike at most once, are considerably more energy efficient than standard artificial neural networks (ANNs). However, single-spike SSNs are difficult to train due to their dynamic and non-differentiable nature, where current solutions are either slow or suffer from training instabilities. These networks have also been critiqued for their limited computational applicability such as being unsuitable for time-series datasets. We propose a new model for training single-spike SNNs which mitigates the aforementioned training issues and obtains competitive results across various image and neuromorphic datasets, with up to a $13.98\times$ training speedup and up to an $81\%$ reduction in spikes compared to the multi-spike SNN. Notably, our model performs on par with multi-spike SNNs in challenging tasks involving neuromorphic time-series datasets, demonstrating a broader computational role for single-spike SNNs than previously believed.
Reject
ICLR.cc/2022/Conference
Do deep networks transfer invariances across classes?
In order to generalize well, classifiers must learn to be invariant to nuisance transformations that do not alter an input's class. Many problems have "class-agnostic" nuisance transformations that apply similarly to all classes, such as lighting and background changes for image classification. Neural networks can learn these invariances given sufficient data, but many real-world datasets are heavily class imbalanced and contain only a few examples for most of the classes. We therefore pose the question: how well do neural networks transfer class-agnostic invariances learned from the large classes to the small ones? Through careful experimentation, we observe that invariance to class-agnostic transformations is still heavily dependent on class size, with the networks being much less invariant on smaller classes. This result holds even when using data balancing techniques, and suggests poor invariance transfer across classes. Our results provide one explanation for why classifiers generalize poorly on unbalanced and long-tailed distributions. Based on this analysis, we show how a generative approach for learning the nuisance transformations can help transfer invariances across classes and improve performance on a set of imbalanced image classification benchmarks.
Accept (Poster)
ICLR.cc/2023/Conference
Fast Sampling of Diffusion Models with Exponential Integrator
The past few years have witnessed the great success of Diffusion models~(DMs) in generating high-fidelity samples in generative modeling tasks. A major limitation of the DM is its notoriously slow sampling procedure which normally requires hundreds to thousands of time discretization steps of the learned diffusion process to reach the desired accuracy. Our goal is to develop a fast sampling method for DMs with a much less number of steps while retaining high sample quality. To this end, we systematically analyze the sampling procedure in DMs and identify key factors that affect the sample quality, among which the method of discretization is most crucial. By carefully examining the learned diffusion process, we propose Diffusion Exponential Integrator Sampler~(DEIS). It is based on the Exponential Integrator designed for discretizing ordinary differential equations (ODEs) and leverages a semilinear structure of the learned diffusion process to reduce the discretization error. The proposed method can be applied to any DMs and can generate high-fidelity samples in as few as 10 steps. Moreover, by directly using pre-trained DMs, we achieve state-of-art sampling performance when the number of score function evaluation~(NFE) is limited, e.g., 4.17 FID with 10 NFEs, 2.86 FID with only 20 NFEs on CIFAR10.
Accept: poster
ICLR.cc/2020/Conference
From English to Foreign Languages: Transferring Pre-trained Language Models
Pre-trained models have demonstrated their effectiveness in many downstream natural language processing (NLP) tasks. The availability of multilingual pre-trained models enables zero-shot transfer of NLP tasks from high resource languages to low resource ones. However, recent research in improving pre-trained models focuses heavily on English. While it is possible to train the latest neural architectures for other languages from scratch, it is undesirable due to the required amount of compute. In this work, we tackle the problem of transferring an existing pre-trained model from English to other languages under a limited computational budget. With a single GPU, our approach can obtain a foreign BERT-base model within a day and a foreign BERT-large within two days. Furthermore, evaluating our models on six languages, we demonstrate that our models are better than multilingual BERT on two zero-shot tasks: natural language inference and dependency parsing.
Reject
ICLR.cc/2021/Conference
Representation and Bias in Multilingual NLP: Insights from Controlled Experiments on Conditional Language Modeling
Inspired by the phenomenon of performance disparity between languages in machine translation, we investigate whether and to what extent languages are equally hard to "conditional-language-model". Our goal is to improve our understanding and expectation of the relationship between language, data representation, size, and performance. We study one-to-one, bilingual conditional language modeling through a series of systematically controlled experiments with the Transformer and the 6 languages from the United Nations Parallel Corpus. We examine character, byte, and word models in 30 language directions and 5 data sizes, and observe indications suggesting a script bias on the character level, a length bias on the byte level, and a word bias that gives rise to a hierarchy in performance across languages. We also identify two types of sample-wise non-monotonicity --- while word-based representations are prone to exhibit Double Descent, length can induce unstable performance across the size range studied in a novel meta phenomenon which we term "erraticity". By eliminating statistically significant performance disparity on the character and byte levels by normalizing length and vocabulary in the data, we show that, in the context of computing with the Transformer, there is no complexity intrinsic to languages other than that related to their statistical attributes and that performance disparity is not a necessary condition but a byproduct of word segmentation. Our application of statistical comparisons as a fairness measure also serves as a novel rigorous method for the intrinsic evaluation of languages, resolving a decades-long debate on language complexity. While these quantitative biases leading to disparity are mitigable through a shallower network, we find room for a human bias to be reflected upon. We hope our work helps open up new directions in the area of language and computing that would be fairer and more flexible and foster a new transdisciplinary perspective for DL-inspired scientific progress.
Reject
ICLR.cc/2021/Conference
Matrix Data Deep Decoder - Geometric Learning for Structured Data Completion
In this work, we present a fully convolutional end to end method to reconstruct corrupted sparse matrices of Non-Euclidean data. The classic example for such matrices is recommender systems matrices where the rows/columns represent items/users and the entries are ratings. The method we present is inspired by the surprising and spectacular success of methods like$"$ deep image prior$"$ and $``$deep decoder$"$ for corrupted image completion. In sharp contrast to previous Matrix Completion methods wherein the latent matrix or its factors directly serve as the optimization variable, in the method we present, the matrix is parameterized as the weights of a graph neural network acting on a random noisy input. Then we are tuning the network parameters to get a result as close as possible to the initial sparse matrix (using its factors) getting that way state of the art matrix completion result. In addition to the conceptual simplicity of our method, which is just Non-Euclidean generalization of deep image priors, it holds fewer parameters than previously presented methods which makes the parameters more trackable and the method more computationally efficient and more applicable for the real-world tasks. The method also achieves state-of-the-art results for the matrix completion task on the classical benchmarks in the field. The method also surprisingly shows that untrained convolutional neural network can use a good prior not only for image completion but also for Matrix Completion when redefined for graphs.
Reject
ICLR.cc/2023/Conference
Model-agnostic Measure of Generalization Difficulty
The measure of a machine learning algorithm is the difficulty of the tasks it can perform, and sufficiently difficult tasks are critical drivers of strong machine learning models. However, quantifying the generalization difficulty of machine learning benchmarks has remained challenging. We propose what is to our knowledge the first model-agnostic measure of the inherent generalization difficulty of tasks. Our inductive bias complexity measure quantifies the total information required to generalize well on a task minus the information provided by the data. It does so by measuring the fractional volume occupied by hypotheses that generalize on a task given that they fit the training data. It scales exponentially with the intrinsic dimensionality of the space over which the model must generalize but only polynomially in resolution per dimension, showing that tasks which require generalizing over many dimensions are drastically more difficult than tasks involving more detail in fewer dimensions. Our measure can be applied to compute and compare supervised learning, reinforcement learning and meta-learning task difficulties against each other. We show that applied empirically, it formally quantifies intuitively expected trends, e.g. that in terms of required inductive bias, MNIST $<$ CIFAR10 $<$ Imagenet and fully observable Markov decision processes (MDPs) $<$ partially observable MDPs. Further, we show that classification of complex images $<$ few-shot meta-learning with simple images. Our measure provides a quantitative metric to guide the construction of more complex tasks requiring greater inductive bias, and thereby encourages the development of more sophisticated architectures and learning algorithms with more powerful generalization capabilities.
Reject
ICLR.cc/2018/Conference
GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks
Deep multitask networks, in which one neural network produces multiple predictive outputs, are more scalable and often better regularized than their single-task counterparts. Such advantages can potentially lead to gains in both speed and performance, but multitask networks are also difficult to train without finding the right balance between tasks. We present a novel gradient normalization (GradNorm) technique which automatically balances the multitask loss function by directly tuning the gradients to equalize task training rates. We show that for various network architectures, for both regression and classification tasks, and on both synthetic and real datasets, GradNorm improves accuracy and reduces overfitting over single networks, static baselines, and other adaptive multitask loss balancing techniques. GradNorm also matches or surpasses the performance of exhaustive grid search methods, despite only involving a single asymmetry hyperparameter $\alpha$. Thus, what was once a tedious search process which incurred exponentially more compute for each task added can now be accomplished within a few training runs, irrespective of the number of tasks. Ultimately, we hope to demonstrate that gradient manipulation affords us great control over the training dynamics of multitask networks and may be one of the keys to unlocking the potential of multitask learning.
Reject
ICLR.cc/2022/Conference
Text Style Transfer with Confounders
Existing methods for style transfer operate either with paired sentences or distributionally matched corpora which differ only in the desired style. In this paper, we relax this restriction and consider data sources with additional confounding differences, from which the desired style needs to be inferred. Specifically, we first learn an invariant style classifier that takes out nuisance variation, and then introduce an orthogonal classifier that highlights the confounding cues. The resulting pair of classifiers guide us to transfer text in the specified direction, creating sentences of the type not seen during training. Experiments show that using positive and negative review datasets from different categories, we can successfully transfer the sentiment without changing the category.
Reject
ICLR.cc/2019/Conference
The GAN Landscape: Losses, Architectures, Regularization, and Normalization
Generative adversarial networks (GANs) are a class of deep generative models which aim to learn a target distribution in an unsupervised fashion. While they were successfully applied to many problems, training a GAN is a notoriously challenging task and requires a significant amount of hyperparameter tuning, neural architecture engineering, and a non-trivial amount of ``tricks". The success in many practical applications coupled with the lack of a measure to quantify the failure modes of GANs resulted in a plethora of proposed losses, regularization and normalization schemes, and neural architectures. In this work we take a sober view of the current state of GANs from a practical perspective. We reproduce the current state of the art and go beyond fairly exploring the GAN landscape. We discuss common pitfalls and reproducibility issues, open-source our code on Github, and provide pre-trained models on TensorFlow Hub.
Reject
ICLR.cc/2023/Conference
Peaks2Image: Reconstructing fMRI Statistical Maps from Peaks
Neuroscience is striving to overcome the lack of power due to the small sample size of standard studies. An important step forward has been the creation of large-scale public image repositories, such as NeuroVault. Such repositories enable comparing images across studies and automatically associating them with cognitive terms. Yet, this type of meta-analysis faces a major roadblock: the scarcity and inconsistency of image annotations and metadata. Another resource containing rich annotations is the neuroscientific literature. However it only yields a handful of brain-space coordinates per publication, those of the main activity peaks reported in each study. In this work, we propose Peaks2Image, a neuralnetwork approach to reconstructing continuous spatial representations of brain activity from peak activation tables. Using reconstructions of studies published in the neuroscientific literature, we train a decoder using tf-idf features as labels, leading to a much broader set of decoded terms than current image-based studies. We validate the decoder on 43,000 NeuroVault images, successfully decoding 58 out of 81 concepts in a zero-shot setting.
Reject
ICLR.cc/2020/Conference
CAPACITY-LIMITED REINFORCEMENT LEARNING: APPLICATIONS IN DEEP ACTOR-CRITIC METHODS FOR CONTINUOUS CONTROL
Biological and artificial agents must learn to act optimally in spite of a limited capacity for processing, storing, and attending to information. We formalize this type of bounded rationality in terms of an information-theoretic constraint on the complexity of policies that agents seek to learn. We present the Capacity-Limited Reinforcement Learning (CLRL) objective which defines an optimal policy subject to an information capacity constraint. This objective is optimized by drawing from methods used in rate distortion theory and information theory, and applied to the reinforcement learning setting. Using this objective we implement a novel Capacity-Limited Actor-Critic (CLAC) algorithm and situate it within a broader family of RL algorithms such as the Soft Actor Critic (SAC) and discuss their similarities and differences. Our experiments show that compared to alternative approaches, CLAC offers improvements in generalization between training and modified test environments. This is achieved in the CLAC model while displaying high sample efficiency and minimal requirements for hyper-parameter tuning.
Reject
ICLR.cc/2020/Conference
Learning to Remember from a Multi-Task Teacher
Recent studies on catastrophic forgetting during sequential learning typically focus on fixing the accuracy of the predictions for a previously learned task. In this paper we argue that the outputs of neural networks are subject to rapid changes when learning a new data distribution, and networks that appear to "forget" everything still contain useful representation towards previous tasks. We thus propose to enforce the output accuracy to stay the same, we should aim to reduce the effect of catastrophic forgetting on the representation level, as the output layer can be quickly recovered later with a small number of examples. Towards this goal, we propose an experimental setup that measures the amount of representational forgetting, and develop a novel meta-learning algorithm to overcome this issue. The proposed meta-learner produces weight updates of a sequential learning network, mimicking a multi-task teacher network's representation. We show that our meta-learner can improve its learned representations on new tasks, while maintaining a good representation for old tasks.
Reject
ICLR.cc/2023/Conference
Task Regularized Hybrid Knowledge Distillation For Continual Object Detection
Knowledge distillation has been used to overcome catastrophic forgetting in Continual Object Detection(COD) task. Previous works mainly focus on combining different distillation methods, including feature, classification, location and relation, into a mixed scheme to solve this problem. In this paper, we propose a task regularized hybrid knowledge distillation method for COD task. First, we propose an image-level hybrid knowledge representation by combining instance-level hard and soft knowledge to use teacher knowledge critically. Second, we propose a task-based regularization distillation loss by taking account of loss and category differences to make continual learning more balance between old and new tasks. We find that, under appropriate knowledge selection and transfer strategies, using only classification distillation can also relieve knowledge forgetting effectively. Extensive experiments conducted on MS COCO2017 demonstrate that our method achieves state-of-the-art results under various scenarios. We get an absolute improvement of 27.98 at $RelGap$ under the most difficult five-task scenario. Code is in attachment and will be available on github.
Reject
ICLR.cc/2023/Conference
AutoShot: A Short Video Dataset and State-of-the-Art Shot Boundary Detection
The short-form videos have explosive popularity and have dominated the new social media trends. Prevailing short-video platforms, e.g., TikTok, Instagram Reels, and YouTube Shorts, have changed the way we consume and create content. For video content creation and understanding, the shot boundary detection (SBD) is one of the most essential components in various scenarios. In this work, we release a new public Short video sHot bOundary deTection dataset, named SHOT, consisting of 853 complete short videos and 11,606 shot annotations, with 2,716 high quality shot boundary annotations in 200 test videos. Leveraging this new data wealth, we propose to optimize the model design for video SBD, by conducting neural architecture search in a search space encapsulating various advanced 3D ConvNets and Transformers. Our proposed approach, named AutoShot, achieves higher F1 scores than previous state-of-the-art approaches, e.g., outperforming TransNetV2 by 4.2%, when being derived and evaluated on our newly constructed SHOT dataset. Moreover, to validate the generalizability of the AutoShot architecture, we directly evaluate it on another three public datasets: ClipShots, BBC and RAI, and the F1 scores of AutoShot outperform previous state-of-the-art approaches by 1.1%, 0.9% and 1.2%, respectively. The SHOT dataset and code will be released.
Reject
ICLR.cc/2023/Conference
Data augmentation alone can improve adversarial training
Adversarial training suffers from the issue of robust overfitting, which seriously impairs its generalization performance. Data augmentation, which is effective at preventing overfitting in standard training, has been observed by many previous works to be ineffective in mitigating overfitting in adversarial training. This work proves that, contrary to previous findings, data augmentation alone can significantly boost accuracy and robustness in adversarial training. We find that the hardness and the diversity of data augmentation are important factors in combating robust overfitting. In general, diversity can improve both accuracy and robustness, while hardness can boost robustness at the cost of accuracy within a certain limit and degrade them both over that limit. To mitigate robust overfitting, we first propose a new crop transformation Cropshift with improved diversity compared to the conventional one (Padcrop). We then propose a new data augmentation scheme, based on Cropshift, with much improved diversity and well-balanced hardness. Empirically, our augmentation method achieves the state-of-the-art accuracy and robustness for data augmentations in adversarial training. Furthermore, it matches, or even exceeds when combined with weight averaging, the performance of the best contemporary regularization methods for alleviating robust overfitting.
Accept: poster
ICLR.cc/2022/Conference
CARD: Certifiably Robust Machine Learning Pipeline via Domain Knowledge Integration
The advent of ubiquitous machine learning (ML) has led to exciting revolution in computing today. However, recent studies have shown that ML, especially deep neural networks (DNNs), are vulnerable to adversarial examples, which are able to mislead DNNs with carefully crafted stealthy perturbations. So far, many defense approaches have been proposed against such adversarial attacks, both empirically and theoretically. Though effective under certain conditions, existing empirical defenses are usually found vulnerable against new attacks; existing certified defenses are only able to certify robustness against limited perturbation radius. As current pure data-driven defenses have reached a bottleneck towards certifiably robust ML, in this paper we propose a certifiably robust ML pipeline CARD, aiming to integrate exogenous information, such as domain knowledge, as logical rules with ML models to improve the certified robustness. Intuitively, domain knowledge (e.g., cat belongs to the animal category) will prevent attacks that violate these knowledge rules, and it is also challenging to construct adaptive attacks satisfying such pre-defined logical relationships. In particular, we express the domain knowledge as first-order logic rules and embed these logic rules in a probabilistic graphical model. We then prove that such a probabilistic graphical model can be mapped to a 1-layer NN for efficient training. We conduct extensive experiments on several high-dimensional datasets and show that our proposed CARD achieves the state-of-the-art certified robustness.
Withdrawn
ICLR.cc/2020/Conference
BERT for Sequence-to-Sequence Multi-Label Text Classification
We study the BERT language representation model and the sequence generation model with BERT encoder for multi-label text classification task. We experiment with both models and explore their special qualities for this setting. We also introduce and examine experimentally a mixed model, which is an ensemble of multi-label BERT and sequence generating BERT models. Our experiments demonstrated that BERT-based models and the mixed model, in particular, outperform current baselines in several metrics achieving state-of-the-art results on three well-studied multi-label classification datasets with English texts and two private Yandex Taxi datasets with Russian texts.
Withdrawn
ICLR.cc/2021/Conference
Grounding Physical Concepts of Objects and Events Through Dynamic Visual Reasoning
We study the problem of dynamic visual reasoning on raw videos. This is a challenging problem; currently, state-of-the-art models often require dense supervision on physical object properties and events from simulation, which are impractical to obtain in real life. In this paper, we present the Dynamic Concept Learner (DCL), a unified framework that grounds physical objects and events from video and language. DCL first adopts a trajectory extractor to track each object over time and to represent it as a latent, object-centric feature vector. Building upon this object-centric representation, DCL learns to approximate the dynamic interaction among objects using graph networks. DCL further incorporates a semantic parser to parse question into semantic programs and, finally, a program executor to run the program to answer the question, levering the learned dynamics model. After training, DCL can detect and associate objects across the frames, ground visual properties and physical events, understand the causal relationship between events, make future and counterfactual predictions, and leverage these extracted presentations for answering queries. DCL achieves state-of-the-art performance on CLEVRER, a challenging causal video reasoning dataset, even without using ground-truth attributes and collision labels from simulations for training. We further test DCL on a newly proposed video-retrieval and event localization dataset derived from CLEVRER, showing its strong generalization capacity.
Accept (Poster)
ICLR.cc/2021/Conference
Probabilistic Numeric Convolutional Neural Networks
Continuous input signals like images and time series that are irregularly sampled or have missing values are challenging for existing deep learning methods. Coherently defined feature representations must depend on the values in unobserved regions of the input. Drawing from the work in probabilistic numerics, we propose Probabilistic Numeric Convolutional Neural Networks which represent features as Gaussian processes, providing a probabilistic description of discretization error. We then define a convolutional layer as the evolution of a PDE defined on this GP, followed by a nonlinearity. This approach also naturally admits steerable equivariant convolutions under e.g. the rotation group. In experiments we show that our approach yields a $3\times$ reduction of error from the previous state of the art on the SuperPixel-MNIST dataset and competitive performance on the medical time series dataset PhysioNet2012.
Accept (Poster)
ICLR.cc/2020/Conference
Towards Understanding Generalization in Gradient-Based Meta-Learning
In this work we study generalization of neural networks in gradient-based meta-learning by analyzing various properties of the objective landscapes. We experimentally demonstrate that as meta-training progresses, the meta-test solutions obtained by adapting the meta-train solution of the model to new tasks via few steps of gradient-based fine-tuning, become flatter, lower in loss, and further away from the meta-train solution. We also show that those meta-test solutions become flatter even as generalization starts to degrade, thus providing an experimental evidence against the correlation between generalization and flat minima in the paradigm of gradient-based meta-leaning. Furthermore, we provide empirical evidence that generalization to new tasks is correlated with the coherence between their adaptation trajectories in parameter space, measured by the average cosine similarity between task-specific trajectory directions, starting from a same meta-train solution. We also show that coherence of meta-test gradients, measured by the average inner product between the task-specific gradient vectors evaluated at meta-train solution, is also correlated with generalization.
Withdrawn
ICLR.cc/2023/Conference
Collecting The Puzzle Pieces: Disentangled Self-Driven Human Pose Transfer by Permuting Textures
Human pose transfer aims to synthesize a new view of a person under a given pose. Recent works achieve this via self-reconstruction, which disentangles pose and texture features from the person image, then combines the two features to reconstruct the person. Such feature-level disentanglement is a difficult and ill-defined problem that could lead to loss of details and unwanted artifacts. In this paper, we propose a self-driven human pose transfer method that permutes the textures at random, then reconstructs the image with a dual branch attention to achieve image-level disentanglement and detail-preserving texture transfer. We find that compared with feature-level disentanglement, image-level disentanglement is more controllable and reliable. Furthermore, we introduce a dual kernel encoder that gives different sizes of receptive fields in order to reduce the noise caused by permutation and thus recover clothing details while aligning pose and textures. Extensive experiments on DeepFashion and Market-1501 shows that our model improves the quality of generated images in terms of FID, LPIPS and SSIM over other self-driven methods, and even outperforming some fully-supervised methods. A user study also shows that among self-driven approaches, images generated by our method are preferred in 72\% of cases over prior work.
Withdrawn
ICLR.cc/2019/Conference
INVASE: Instance-wise Variable Selection using Neural Networks
The advent of big data brings with it data with more and more dimensions and thus a growing need to be able to efficiently select which features to use for a variety of problems. While global feature selection has been a well-studied problem for quite some time, only recently has the paradigm of instance-wise feature selection been developed. In this paper, we propose a new instance-wise feature selection method, which we term INVASE. INVASE consists of 3 neural networks, a selector network, a predictor network and a baseline network which are used to train the selector network using the actor-critic methodology. Using this methodology, INVASE is capable of flexibly discovering feature subsets of a different size for each instance, which is a key limitation of existing state-of-the-art methods. We demonstrate through a mixture of synthetic and real data experiments that INVASE significantly outperforms state-of-the-art benchmarks.
Accept (Poster)
ICLR.cc/2020/Conference
A Theoretical Analysis of the Number of Shots in Few-Shot Learning
Few-shot classification is the task of predicting the category of an example from a set of few labeled examples. The number of labeled examples per category is called the number of shots (or shot number). Recent works tackle this task through meta-learning, where a meta-learner extracts information from observed tasks during meta-training to quickly adapt to new tasks during meta-testing. In this formulation, the number of shots exploited during meta-training has an impact on the recognition performance at meta-test time. Generally, the shot number used in meta-training should match the one used in meta-testing to obtain the best performance. We introduce a theoretical analysis of the impact of the shot number on Prototypical Networks, a state-of-the-art few-shot classification method. From our analysis, we propose a simple method that is robust to the choice of shot number used during meta-training, which is a crucial hyperparameter. The performance of our model trained for an arbitrary meta-training shot number shows great performance for different values of meta-testing shot numbers. We experimentally demonstrate our approach on different few-shot classification benchmarks.
Accept (Poster)
ICLR.cc/2020/Conference
Reconstructing continuous distributions of 3D protein structure from cryo-EM images
Cryo-electron microscopy (cryo-EM) is a powerful technique for determining the structure of proteins and other macromolecular complexes at near-atomic resolution. In single particle cryo-EM, the central problem is to reconstruct the 3D structure of a macromolecule from $10^{4-7}$ noisy and randomly oriented 2D projection images. However, the imaged protein complexes may exhibit structural variability, which complicates reconstruction and is typically addressed using discrete clustering approaches that fail to capture the full range of protein dynamics. Here, we introduce a novel method for cryo-EM reconstruction that extends naturally to modeling continuous generative factors of structural heterogeneity. This method encodes structures in Fourier space using coordinate-based deep neural networks, and trains these networks from unlabeled 2D cryo-EM images by combining exact inference over image orientation with variational inference for structural heterogeneity. We demonstrate that the proposed method, termed cryoDRGN, can perform ab-initio reconstruction of 3D protein complexes from simulated and real 2D cryo-EM image data. To our knowledge, cryoDRGN is the first neural network-based approach for cryo-EM reconstruction and the first end-to-end method for directly reconstructing continuous ensembles of protein structures from cryo-EM images.
Accept (Spotlight)
ICLR.cc/2022/Conference
Message Passing Neural PDE Solvers
The numerical solution of partial differential equations (PDEs) is difficult, having led to a century of research so far. Recently, there have been pushes to build neural--numerical hybrid solvers, which piggy-backs the modern trend towards fully end-to-end learned systems. Most works so far can only generalize over a subset of properties to which a generic solver would be faced, including: resolution, topology, geometry, boundary conditions, domain discretization regularity, dimensionality, etc. In this work, we build a solver, satisfying these properties, where all the components are based on neural message passing, replacing all heuristically designed components in the computation graph with backprop-optimized neural function approximators. We show that neural message passing solvers representationally contain some classical methods, such as finite differences, finite volumes, and WENO schemes. In order to encourage stability in training autoregressive models, we put forward a method that is based on the principle of zero-stability, posing stability as a domain adaptation problem. We validate our method on various fluid-like flow problems, demonstrating fast, stable, and accurate performance across different domain topologies, discretization, etc. in 1D and 2D. Our model outperforms state-of-the-art numerical solvers in the low resolution regime in terms of speed, and accuracy.
Accept (Spotlight)
ICLR.cc/2020/Conference
Fix-Net: pure fixed-point representation of deep neural networks
Deep neural networks (DNNs) dominate current research in machine learning. Due to massive GPU parallelization DNN training is no longer a bottleneck, and large models with many parameters and high computational effort lead common benchmark tables. In contrast, embedded devices have a very limited capability. As a result, both model size and inference time must be significantly reduced if DNNs are to achieve suitable performance on embedded devices. We propose a soft quantization approach to train DNNs that can be evaluated using pure fixed-point arithmetic. By exploiting the bit-shift mechanism, we derive fixed-point quantization constraints for all important components, including batch normalization and ReLU. Compared to floating-point arithmetic, fixed-point calculations significantly reduce computational effort whereas low-bit representations immediately decrease memory costs. We evaluate our approach with different architectures on common benchmark data sets and compare with recent quantization approaches. We achieve new state of the art performance using 4-bit fixed-point models with an error rate of 4.98% on CIFAR-10.
Withdrawn
ICLR.cc/2022/Conference
Predicting Unreliable Predictions by Shattering a Neural Network
Generalization error bounds measure the deviation of performance on unseen test data from performance on training data. However, by providing one scalar per model, they are input-agnostic. What if one wants to predict error for a specific test sample? To answer this, we propose the novel paradigm of input-conditioned generalization error bounds. For piecewise linear neural networks, given a weighting function that relates the errors of different input activation regions together, we obtain a bound on each region's generalization error that scales inversely with the density of training samples. That is, more densely supported regions are more reliable. As the bound is input-conditioned, it is to our knowledge the first generalization error bound applicable to the problems of detecting out-of-distribution and misclassified in-distribution samples for neural networks; we find that it performs competitively in both cases when tested on image classification tasks.
Withdrawn
ICLR.cc/2022/Conference
Learning to Map for Active Semantic Goal Navigation
We consider the problem of object goal navigation in unseen environments. Solving this problem requires learning of contextual semantic priors, a challenging endeavour given the spatial and semantic variability of indoor environments. Current methods learn to implicitly encode these priors through goal-oriented navigation policy functions operating on spatial representations that are limited to the agent's observable areas. In this work, we propose a novel framework that actively learns to generate semantic maps outside the field of view of the agent and leverages the uncertainty over the semantic classes in the unobserved areas to decide on long term goals. We demonstrate that through this spatial prediction strategy, we are able to learn semantic priors in scenes that can be leveraged in unknown environments. Additionally, we show how different objectives can be defined by balancing exploration with exploitation during searching for semantic targets. Our method is validated in the visually realistic environments of the Matterport3D dataset and show improved results on object goal navigation over competitive baselines.
Accept (Poster)
ICLR.cc/2023/Conference
Differentiable Rendering with Reparameterized Volume Sampling
We propose an alternative rendering algorithm for neural radiance fields based on importance sampling. In view synthesis, a neural radiance field approximates underlying density and radiance fields based on a sparse set of views of a scene. To generate a pixel of a novel view, it marches a ray through the pixel and computes a weighted sum of radiance emitted from a dense set of ray points. This rendering algorithm is fully differentiable and facilitates gradient-based optimization of the fields. However, in practice, only a tiny opaque portion of the ray contributes most of the radiance to the sum. Therefore, we can avoid computing radiance in the rest part. In this work, we use importance sampling to pick non-transparent points on the ray. Specifically, we generate samples according to the probability distribution induced by the density field. Our main contribution is the reparameterization of the sampling algorithm. It allows end-to-end learning with gradient descent as in the original rendering algorithm. With our approach, we can optimize a neural radiance field with just a few radiance field evaluations per ray. As a result, we alleviate the costs associated with the color component of the neural radiance field.
Reject
ICLR.cc/2020/Conference
Connecting the Dots Between MLE and RL for Sequence Prediction
Sequence prediction models can be learned from example sequences with a variety of training algorithms. Maximum likelihood learning is simple and efficient, yet can suffer from compounding error at test time. Reinforcement learning such as policy gradient addresses the issue but can have prohibitively poor exploration efficiency. A rich set of other algorithms, such as data noising, RAML, and softmax policy gradient, have also been developed from different perspectives. In this paper, we present a formalism of entropy regularized policy optimization, and show that the apparently distinct algorithms, including MLE, can be reformulated as special instances of the formulation. The difference between them is characterized by the reward function and two weight hyperparameters. The unifying interpretation enables us to systematically compare the algorithms side-by-side, and gain new insights into the trade-offs of the algorithm design. The new perspective also leads to an improved approach that dynamically interpolates among the family of algorithms, and learns the model in a scheduled way. Experiments on machine translation, text summarization, and game imitation learning demonstrate superiority of the proposed approach.
Reject
ICLR.cc/2021/Conference
Trajectory Prediction using Equivariant Continuous Convolution
Trajectory prediction is a critical part of many AI applications, for example, the safe operation of autonomous vehicles. However, current methods are prone to making inconsistent and physically unrealistic predictions. We leverage insights from fluid dynamics to overcome this limitation by considering internal symmetry in real-world trajectories. We propose a novel model, Equivariant Continous COnvolution (ECCO) for improved trajectory prediction. ECCO uses rotationally-equivariant continuous convolutions to embed the symmetries of the system. On both vehicle and pedestrian trajectory datasets, ECCO attains competitive accuracy with significantly fewer parameters. It is also more sample efficient, generalizing automatically from few data points in any orientation. Lastly, ECCO improves generalization with equivariance, resulting in more physically consistent predictions. Our method provides a fresh perspective towards increasing trust and transparency in deep learning models. Our code and data can be found at https://github.com/Rose-STL-Lab/ECCO.
Accept (Poster)
ICLR.cc/2022/Conference
Adaptive Region Pooling for Fine-Grained Representation Learning
Fine-grained recognition aims to discriminate the sub-categories of the images within one general category. It is fundamentally difficult due to the requirement to extract fine-grained features from subtle regions. Nonetheless, a Convolutional Neural Network typically applies strided operations to downsample the representation, which would excessively spoil the feature resolution and lead to a significant loss of fine-grained information. In this paper, we propose Adaptive Region Pooling (ARP): a novel downsampling algorithm that makes the network only focus on a smaller but more critical region, and simultaneously increase the resolution of sub-sampled feature. ARP owns a trade-off mechanism that allows users to actively balance the scale of receptive field and the granularity of feature. Also, without any learning-based parameters, ARP provides the network a stabler training process and an earlier convergence. Extensive experiments qualitatively and quantitatively validate the effectiveness and efficiency of the proposed pooling operation and show superior performance against the state-of-the-arts in both the tasks of image classification and image retrieval.
Withdrawn
ICLR.cc/2021/Conference
Representational aspects of depth and conditioning in normalizing flows
Normalizing flows are among the most popular paradigms in generative modeling, especially for images, primarily because we can efficiently evaluate the likelihood of a data point. This is desirable both for evaluating the fit of a model, and for ease of training, as maximizing the likelihood can be done by gradient descent. However, training normalizing flows comes with difficulties as well: models which produce good samples typically need to be extremely deep -- which comes with accompanying vanishing/exploding gradient problems. A very related problem is that they are often poorly \emph{conditioned}: since they are parametrized as invertible maps from $\mathbb{R}^d \to \mathbb{R}^d$, and typical training data like images intuitively is lower-dimensional, the learned maps often have Jacobians that are close to being singular. In our paper, we tackle representational aspects around depth and conditioning of normalizing flows---both for general invertible architectures, and for a particular common architecture---affine couplings. For general invertible architectures, we prove that invertibility comes at a cost in terms of depth: we show examples where a much deeper normalizing flow model may need to be used to match the performance of a non-invertible generator. For affine couplings, we first show that the choice of partitions isn't a likely bottleneck for depth: we show that any invertible linear map (and hence a permutation) can be simulated by a constant number of affine coupling layers, using a fixed partition. This shows that the extra flexibility conferred by 1x1 convolution layers, as in GLOW, can in principle be simulated by increasing the size by a constant factor. Next, in terms of conditioning, we show that affine couplings are universal approximators -- provided the Jacobian of the model is allowed to be close to singular. We furthermore empirically explore the benefit of different kinds of padding -- a common strategy for improving conditioning.
Reject
ICLR.cc/2021/Conference
Adversarial Attacks on Machine Learning Systems for High-Frequency Trading
Algorithmic trading systems are often completely automated, and deep learning is increasingly receiving attention in this domain. Nonetheless, little is known about the robustness properties of these models. We study valuation models for algorithmic trading from the perspective of adversarial machine learning. We introduce new attacks specific to this domain with size constraints that minimize attack costs. We further discuss how these attacks can be used as an analysis tool to study and evaluate the robustness properties of financial models. Finally, we investigate the feasibility of realistic adversarial attacks in which an adversarial trader fools automated trading systems into making inaccurate predictions.
Withdrawn