File size: 6,591 Bytes
ee3928c 8993d83 452d29b cd5873c ee3928c 833aeeb bb0f400 66a1c60 13094c8 66a1c60 833aeeb 8993d83 738abf4 3ffbf1b 8993d83 738abf4 8993d83 89de5d4 17fc37d 89de5d4 17fc37d 738abf4 8993d83 89de5d4 17fc37d 89de5d4 17fc37d 89de5d4 17fc37d 89de5d4 17fc37d 738abf4 17fc37d 833aeeb 8d9bf15 738abf4 8993d83 837b74b cd5873c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 |
---
license: apache-2.0
datasets:
- anon8231489123/ShareGPT_Vicuna_unfiltered
- declare-lab/HarmfulQA
model-index:
- name: starling-7B
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 51.02
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=declare-lab/starling-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 76.77
name: normalized accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=declare-lab/starling-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 47.75
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=declare-lab/starling-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 48.18
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=declare-lab/starling-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 70.56
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=declare-lab/starling-7B
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 10.08
name: accuracy
source:
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=declare-lab/starling-7B
name: Open LLM Leaderboard
---
[**Paper**](https://arxiv.org/abs/2308.09662) | [**Github**](https://github.com/declare-lab/red-instruct) | [**Dataset**](https://huggingface.co/datasets/declare-lab/HarmfulQA)| [**Model**](https://huggingface.co/declare-lab/starling-7B)
> 📣 Update 2/02/24: Introducing Resta: **Safety Re-alignment of Language Models**. [**Paper**](https://arxiv.org/abs/2402.11746) [**Github**](https://github.com/declare-lab/resta) [**Dataset**](https://huggingface.co/datasets/declare-lab/CategoricalHarmfulQ)
As a part of our research efforts to make LLMs safer, we created **Starling**. It is obtained by fine-tuning Vicuna-7B on [**HarmfulQA**](https://huggingface.co/datasets/declare-lab/HarmfulQA), a ChatGPT-distilled dataset that we collected using the Chain of Utterances (CoU) prompt. More details are in our paper [**Red-Teaming Large Language Models using Chain of Utterances for Safety-Alignment**](https://arxiv.org/abs/2308.09662)
<img src="https://declare-lab.github.io/assets/images/logos/starling-final.png" alt="Image" width="100" height="100">
Experimental results on several safety benchmark datasets indicate that **Starling** is a safer model compared to the baseline model, Vicuna.
<img src="https://declare-lab.github.io/assets/images/logos/method.png" alt="Image" width="1000" height="335">
<h2>Experimental Results</h2>
Compared to Vicuna, **Avg. 5.2% reduction in Attack Success Rate** (ASR) on DangerousQA and HarmfulQA using three different prompts.**
Compared to Vicuna, **Avg. 3-7% improvement in HHH score** measured on BBH-HHH benchmark.**
<img src="https://declare-lab.github.io/assets/images/logos/starling-results.png" alt="Image" width="1000" height="335">
TruthfulQA (MC2): **48.90 vs Vicuna's 47.00**
MMLU (5-shot): **46.69 vs Vicuna's 47.18**
BBH (3-shot): **33.47 vs Vicuna's 33.05**
<h2>Jailbreak Prompt for harmfulness eval using Red Eval as reported in the paper</h2>
This jailbreak prompt (termed as Chain of Utterances (CoU) prompt in the paper) shows a 65% Attack Success Rate (ASR) on GPT-4 and 72% on ChatGPT.
<img src="https://declare-lab.github.io/assets/images/logos/jailbreakprompt_main_paper.png" alt="Image" width="1000" height="1000">
<h2>HarmfulQA Data Collection</h2>
We also release our **HarmfulQA** dataset with 1,960 harmful questions (converting 10 topics-10 subtopics) for red-teaming as well as conversations based on them used in model safety alignment, more details [**here**](https://huggingface.co/datasets/declare-lab/HarmfulQA). The following figure describes the data collection process.
<img src="https://declare-lab.github.io/assets/images/logos/data_gen.png" alt="Image" width="1000" height="1000">
_Note: This model is referred to as Starling (Blue) in the paper. We shall soon release Starling (Blue-Red) which was trained on harmful data using an objective function that helps the model learn from the red (harmful) response data._
## Citation
```bibtex
@misc{bhardwaj2023redteaming,
title={Red-Teaming Large Language Models using Chain of Utterances for Safety-Alignment},
author={Rishabh Bhardwaj and Soujanya Poria},
year={2023},
eprint={2308.09662},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_declare-lab__starling-7B)
| Metric |Value|
|---------------------------------|----:|
|Avg. |50.73|
|AI2 Reasoning Challenge (25-Shot)|51.02|
|HellaSwag (10-Shot) |76.77|
|MMLU (5-Shot) |47.75|
|TruthfulQA (0-shot) |48.18|
|Winogrande (5-shot) |70.56|
|GSM8k (5-shot) |10.08|
|