# Jailbreak Attack on JailbreakBench (ASR, BERTScore)

Source: [Jailbreak Attack on JailbreakBench (ASR, BERTScore)](<https://www.sota2.com/research/sota/jailbreak-attack-on-jailbreakbench-asr-bertscore>)

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

Data updated: 2026-03-09T07:31:14.087Z

Evaluation of model performance against jailbreak attacks using the JailbreakBench dataset, measured by Attack Success Rate (ASR) and BERTScore.

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 | Jailbreak Attack |
| Dataset (including any stated split) | JailbreakBench |
| Metric | ASR |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 27 of 27 |
| 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/jailbreak-attack-on-jailbreakbench-asr-bertscore.md?metric=ASR>): Higher is better; unit: PERCENT.

- [BERTScore](<https://www.sota2.com/research/sota/jailbreak-attack-on-jailbreakbench-asr-bertscore.md?metric=BERTScore>): Higher is better; unit: SCALAR.

## Recorded results

| Method | ASR | Result date | Report paper |
| --- | --- | --- | --- |
| SAHA | 91 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| SAHA | 85 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| SCAV | 85 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| SAHA | 82 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| SCAV | 71 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| AutoDAN-Turbo | 63 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| AutoDAN | 60 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| ConVA | 59 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| PAIR | 57 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| AutoDAN | 57 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| SCAV | 55 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| GCG | 47 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| AutoDAN | 40 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| CAA | 39 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| GCG | 37 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| CAA | 37 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| PAIR | 30 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| Based | 29 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| ConVA | 19 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| CAA | 14 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| AutoDAN-Turbo | 13 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| AutoDAN-Turbo | 13 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| GCG | 10 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| PAIR | 10 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| Based | 6 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| Based | 6 | Not recorded | 69ae53fb2d603e6de3e42ad3 |
| ConVA | 3 | Not recorded | 69ae53fb2d603e6de3e42ad3 |

### SAHA

ASR: 91. Result ID: 69ae77336dbb7075b9563a5d.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### SAHA

ASR: 85. Result ID: 69ae77336dbb7075b9563a4b.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

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

### SCAV

ASR: 85. Result ID: 69ae77336dbb7075b9563a5a.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75ff1e209131cea6ad3a.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### SAHA

ASR: 82. Result ID: 69ae77336dbb7075b9563a54.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### SCAV

ASR: 71. Result ID: 69ae77336dbb7075b9563a51.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75ff1e209131cea6ad3a.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### AutoDAN-Turbo

ASR: 63. Result ID: 69ae77336dbb7075b9563a50.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77371.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### AutoDAN

ASR: 60. Result ID: 69ae77336dbb7075b9563a58.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

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

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### ConVA

ASR: 59. Result ID: 69ae77336dbb7075b9563a53.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a111e209131cea80eca.md>)

[Implementation](<https://github.com/hr-jin/ConVA>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### PAIR

ASR: 57. Result ID: 69ae77336dbb7075b9563a44.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74311e209131cea60579.md>)

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

### AutoDAN

ASR: 57. Result ID: 69ae77336dbb7075b9563a4f.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

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

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### SCAV

ASR: 55. Result ID: 69ae77336dbb7075b9563a48.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75ff1e209131cea6ad3a.md>)

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

### GCG

ASR: 47. Result ID: 69ae77336dbb7075b9563a4e.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c3b.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### AutoDAN

ASR: 40. Result ID: 69ae77336dbb7075b9563a46.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

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

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

### CAA

ASR: 39. Result ID: 69ae77336dbb7075b9563a52.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74051e209131cea5f4ad.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### GCG

ASR: 37. Result ID: 69ae77336dbb7075b9563a57.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c3b.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### CAA

ASR: 37. Result ID: 69ae77336dbb7075b9563a5b.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74051e209131cea5f4ad.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### PAIR

ASR: 30. Result ID: 69ae77336dbb7075b9563a56.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74311e209131cea60579.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### Based

ASR: 29. Result ID: 69ae77336dbb7075b9563a55.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### ConVA

ASR: 19. Result ID: 69ae77336dbb7075b9563a4a.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a111e209131cea80eca.md>)

[Implementation](<https://github.com/hr-jin/ConVA>)

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

### CAA

ASR: 14. Result ID: 69ae77336dbb7075b9563a49.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74051e209131cea5f4ad.md>)

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

### AutoDAN-Turbo

ASR: 13. Result ID: 69ae77336dbb7075b9563a47.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77371.md>)

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

### AutoDAN-Turbo

ASR: 13. Result ID: 69ae77336dbb7075b9563a59.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78351e209131cea77371.md>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

### GCG

ASR: 10. Result ID: 69ae77336dbb7075b9563a45.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71861e209131cea51c3b.md>)

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

### PAIR

ASR: 10. Result ID: 69ae77336dbb7075b9563a4d.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74311e209131cea60579.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### Based

ASR: 6. Result ID: 69ae77336dbb7075b9563a43.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

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

### Based

ASR: 6. Result ID: 69ae77336dbb7075b9563a4c.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

Recorded evaluation configuration: {"Target Model":"Qwen1.5-7B-Chat"}. This record does not assert independent reproduction.

### ConVA

ASR: 3. Result ID: 69ae77336dbb7075b9563a5c.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53fb2d603e6de3e42ad3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a111e209131cea80eca.md>)

[Implementation](<https://github.com/hr-jin/ConVA>)

Recorded evaluation configuration: {"Target Model":"Deepseek-llm-7B-Chat"}. This record does not assert independent reproduction.

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