# Alignment Defense on AGNEWS

Source: [Alignment Defense on AGNEWS](<https://www.sota2.com/research/sota/alignment-defense-on-agnews>)

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

Data updated: 2026-03-04T09:44:54.793Z

Evaluation of model alignment and defense mechanisms against adversarial fine-tuning techniques using the AG News dataset.

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 | Alignment defense against harmful fine-tuning |
| Dataset (including any stated split) | AGNEWS |
| Metric | HS |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [HS](<https://www.sota2.com/research/sota/alignment-defense-on-agnews.md?metric=HS>): Higher is better; unit: PERCENT.

- [FA](<https://www.sota2.com/research/sota/alignment-defense-on-agnews.md?metric=FA>): Higher is better; unit: PERCENT.

## Recorded results

| Method | HS | Result date | Report paper |
| --- | --- | --- | --- |
| Vaccine | 51.64 | Not recorded | 69a66bc82d603e6de3e3cb4e |
| SFT | 39.91 | Not recorded | 69a66bc82d603e6de3e3cb4e |
| Lisa | 33.33 | Not recorded | 69a66bc82d603e6de3e3cb4e |
| Booster | 19.22 | Not recorded | 69a66bc82d603e6de3e3cb4e |
| Antibody | 1.91 | Not recorded | 69a66bc82d603e6de3e3cb4e |

### Vaccine

HS: 51.64. Result ID: 69a7ff0c6dbb7075b94d3a76.

[Reporting paper](<https://www.sota2.com/research/paper/69a66bc82d603e6de3e3cb4e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69fa9.md>)

[Implementation](<https://github.com/git-disl/Vaccine>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-9B"}. This record does not assert independent reproduction.

### SFT

HS: 39.91. Result ID: 69a7ff0c6dbb7075b94d3a75.

[Reporting paper](<https://www.sota2.com/research/paper/69a66bc82d603e6de3e3cb4e.md>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-9B"}. This record does not assert independent reproduction.

### Lisa

HS: 33.33. Result ID: 69a7ff0c6dbb7075b94d3a77.

[Reporting paper](<https://www.sota2.com/research/paper/69a66bc82d603e6de3e3cb4e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa756a1e209131cea67bac.md>)

[Implementation](<https://github.com/git-disl/Lisa>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-9B"}. This record does not assert independent reproduction.

### Booster

HS: 19.22. Result ID: 69a7ff0c6dbb7075b94d3a78.

[Reporting paper](<https://www.sota2.com/research/paper/69a66bc82d603e6de3e3cb4e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa777b1e209131cea737ab.md>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-9B"}. This record does not assert independent reproduction.

### Antibody

HS: 1.91. Result ID: 69a7ff0c6dbb7075b94d3a79.

[Reporting paper](<https://www.sota2.com/research/paper/69a66bc82d603e6de3e3cb4e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a66bc82d603e6de3e3cb4e.md>)

Recorded evaluation configuration: {"Backbone":"Gemma-2-9B"}. This record does not assert independent reproduction.

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