# Jailbreaking Attack on AdvBench (ASR + GPT)

Source: [Jailbreaking Attack on AdvBench (ASR + GPT)](<https://www.sota2.com/research/sota/jailbreaking-attack-on-advbench-asr-plus-gpt>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-04T17:14:10.745Z

Evaluation of model susceptibility to jailbreak attacks using the AdvBench dataset, specifically reporting Attack Success Rate (ASR) and ASR measured using GPT-based evaluation.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Jailbreaking Attack |
| Dataset (including any stated split) | AdvBench |
| Metric | ASR |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 28 of 28 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [ASR](<https://www.sota2.com/research/sota/jailbreaking-attack-on-advbench-asr-plus-gpt.md?metric=ASR>): Higher is better; unit: PERCENT.

- [ASR (GPT)](<https://www.sota2.com/research/sota/jailbreaking-attack-on-advbench-asr-plus-gpt.md?metric=ASR%20(GPT)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | ASR | Result date | Report paper |
| --- | --- | --- | --- |
| PiF | 97.7 | Not recorded | 694cc4e953fdd5c36173a68a |
| CHaRS | 92.31 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS-PCT | 92.31 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS | 92.31 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS-PCT | 92.31 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| APAs | 92 | Not recorded | 694cc4e953fdd5c36173a68a |
| DirAbl | 91.35 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS-PCT | 91.35 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| DirAbl | 90.38 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS | 90.38 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| DirAbl | 89.42 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS | 86.54 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS-PCT | 86.54 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| RIPPLE | 86 | Not recorded | 694cc4e953fdd5c36173a68a |
| CHaRS-PCT | 85.58 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS | 85.58 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS | 84.62 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| DirAbl | 83.65 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS-PCT | 83.65 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| DirAbl | 82.69 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS | 80.77 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| CHaRS-PCT | 80.77 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| DirAbl | 79.81 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| DirAbl | 78.85 | Not recorded | 69a7bc6a2d603e6de3e3d308 |
| ReNeLLM | 71.6 | Not recorded | 694cc4e953fdd5c36173a68a |
| LLM-FUZZER | 60 | Not recorded | 694cc4e953fdd5c36173a68a |
| AutoDAN | 26.4 | Not recorded | 694cc4e953fdd5c36173a68a |

### PiF

ASR: 97.7. Result ID: 699941505698665ec4ab6216.

[Reporting paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

Recorded evaluation configuration: {"Target Model":"GPT-4"}. This record does not assert independent reproduction.

### CHaRS

ASR: 92.31. Result ID: 69a868526dbb7075b9502e7c.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Llama3.1-8B-Instruct"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 92.31. Result ID: 69a868526dbb7075b9502e7d.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Llama3.1-8B-Instruct"}. This record does not assert independent reproduction.

### CHaRS

ASR: 92.31. Result ID: 69a868526dbb7075b9502e82.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-3B-Instruct"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 92.31. Result ID: 69a868526dbb7075b9502e83.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-3B-Instruct"}. This record does not assert independent reproduction.

### APAs

ASR: 92. Result ID: 699941505698665ec4ab6210.

[Reporting paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ec1e209131cea53904.md>)

Recorded evaluation configuration: {"Target Model":"GPT-4"}. This record does not assert independent reproduction.

### DirAbl

ASR: 91.35. Result ID: 69a868526dbb7075b9502e7b.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Llama3.1-8B-Instruct"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 91.35. Result ID: 69a868526dbb7075b9502e86.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction.

### DirAbl

ASR: 90.38. Result ID: 69a868526dbb7075b9502e81.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-3B-Instruct"}. This record does not assert independent reproduction.

### CHaRS

ASR: 90.38. Result ID: 69a868526dbb7075b9502e85.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction.

### DirAbl

ASR: 89.42. Result ID: 69a868526dbb7075b9502e84.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction.

### CHaRS

ASR: 86.54. Result ID: 69a868526dbb7075b9502e7f.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Llama3.2-3B-Instruct"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 86.54. Result ID: 69a868526dbb7075b9502e80.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Llama3.2-3B-Instruct"}. This record does not assert independent reproduction.

### RIPPLE

ASR: 86. Result ID: 699941505698665ec4ab6212.

[Reporting paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72241e209131cea54c53.md>)

Recorded evaluation configuration: {"Target Model":"GPT-4"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 85.58. Result ID: 69a868526dbb7075b9502e7a.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Gemma2-9B-Instruct"}. This record does not assert independent reproduction.

### CHaRS

ASR: 85.58. Result ID: 69a868526dbb7075b9502e8b.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-32B-Instruct"}. This record does not assert independent reproduction.

### CHaRS

ASR: 84.62. Result ID: 69a868526dbb7075b9502e79.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Gemma2-9B-Instruct"}. This record does not assert independent reproduction.

### DirAbl

ASR: 83.65. Result ID: 69a868526dbb7075b9502e7e.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Llama3.2-3B-Instruct"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 83.65. Result ID: 69a868526dbb7075b9502e8c.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-32B-Instruct"}. This record does not assert independent reproduction.

### DirAbl

ASR: 82.69. Result ID: 69a868526dbb7075b9502e8a.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-32B-Instruct"}. This record does not assert independent reproduction.

### CHaRS

ASR: 80.77. Result ID: 69a868526dbb7075b9502e88.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-14B-Instruct"}. This record does not assert independent reproduction.

### CHaRS-PCT

ASR: 80.77. Result ID: 69a868526dbb7075b9502e89.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-14B-Instruct"}. This record does not assert independent reproduction.

### DirAbl

ASR: 79.81. Result ID: 69a868526dbb7075b9502e78.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Gemma2-9B-Instruct"}. This record does not assert independent reproduction.

### DirAbl

ASR: 78.85. Result ID: 69a868526dbb7075b9502e87.

[Reporting paper](<https://www.sota2.com/research/paper/69a7bc6a2d603e6de3e3d308.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa757b1e209131cea68414.md>)

[Implementation](<https://github.com/andyrdt/refusal_direction>)

Recorded evaluation configuration: {"Base Model":"Qwen2.5-14B-Instruct"}. This record does not assert independent reproduction.

### ReNeLLM

ASR: 71.6. Result ID: 699941505698665ec4ab6215.

[Reporting paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d97.md>)

Recorded evaluation configuration: {"Target Model":"GPT-4"}. This record does not assert independent reproduction.

### LLM-FUZZER

ASR: 60. Result ID: 699941505698665ec4ab6213.

[Reporting paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

Recorded evaluation configuration: {"Target Model":"GPT-4"}. This record does not assert independent reproduction.

### AutoDAN

ASR: 26.4. Result ID: 699941505698665ec4ab6214.

[Reporting paper](<https://www.sota2.com/research/paper/694cc4e953fdd5c36173a68a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72931e209131cea573c9.md>)

Recorded evaluation configuration: {"Target Model":"GPT-4"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
