# Adversarial Prompt Injection Defense on 500 adversarial samples

Source: [Adversarial Prompt Injection Defense on 500 adversarial samples](<https://www.sota2.com/research/sota/adversarial-prompt-injection-defense-on-500-adversarial-samples>)

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

Data updated: 2026-03-18T08:21:50.221Z

Evaluation of defense mechanisms against adversarial prompt injection attacks using the specific set of 500 adversarial samples, measured by Defense Success Rate.

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 | Adversarial Prompt Injection Defense |
| Dataset (including any stated split) | 500 adversarial samples |
| Metric | Defense Success Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Defense Success Rate](<https://www.sota2.com/research/sota/adversarial-prompt-injection-defense-on-500-adversarial-samples.md?metric=Defense%20Success%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Defense Success Rate | Result date | Report paper |
| --- | --- | --- | --- |
| CoMAI | 100 | Not recorded | 69ba31ae0636e5bf4afc0a16 |
| Single-Agent Baseline | 81.55 | Not recorded | 69ba31ae0636e5bf4afc0a16 |
| AI-Interviewer-Bot v3 | 28 | Not recorded | 69ba31ae0636e5bf4afc0a16 |
| LLM-Interviewer | 0 | Not recorded | 69ba31ae0636e5bf4afc0a16 |

### CoMAI

Defense Success Rate: 100. Result ID: 69ba604247f34ce924a43022.

[Reporting paper](<https://www.sota2.com/research/paper/69ba31ae0636e5bf4afc0a16.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ba31ae0636e5bf4afc0a16.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Single-Agent Baseline

Defense Success Rate: 81.55. Result ID: 69ba604247f34ce924a43023.

[Reporting paper](<https://www.sota2.com/research/paper/69ba31ae0636e5bf4afc0a16.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### AI-Interviewer-Bot v3

Defense Success Rate: 28. Result ID: 69ba604247f34ce924a43025.

[Reporting paper](<https://www.sota2.com/research/paper/69ba31ae0636e5bf4afc0a16.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### LLM-Interviewer

Defense Success Rate: 0. Result ID: 69ba604247f34ce924a43024.

[Reporting paper](<https://www.sota2.com/research/paper/69ba31ae0636e5bf4afc0a16.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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