# Persuasion Evaluation on Anthropic

Source: [Persuasion Evaluation on Anthropic](<https://www.sota2.com/research/sota/persuasion-evaluation-on-anthropic>)

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

Data updated: 2026-03-10T07:30:20.818Z

Evaluation of model performance on assessing argument persuasiveness using the Anthropic dataset, measured specifically by Persuasion gain.

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 | Persuasion |
| Dataset (including any stated split) | Anthropic |
| Metric | Persuasion Gain |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [Persuasion Gain](<https://www.sota2.com/research/sota/persuasion-evaluation-on-anthropic.md?metric=Persuasion%20Gain>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Persuasion Gain | Result date | Report paper |
| --- | --- | --- | --- |
| DeepSeek-R1 | 1.33 | Not recorded | 68fa7bb31e209131cea89ffa |
| Claude 3.7 Sonnet | 1.13 | Not recorded | 68fa7bb31e209131cea89ffa |
| GPT-4o | 0.73 | Not recorded | 68fa7bb31e209131cea89ffa |
| Mistral-7B-Instruct-v0.3 | 0.6 | Not recorded | 68fa7bb31e209131cea89ffa |
| Qwen2.5-7B-Instruct | 0.51 | Not recorded | 68fa7bb31e209131cea89ffa |
| Llama-3.3-70B-Instruct | 0.49 | Not recorded | 68fa7bb31e209131cea89ffa |
| Llama-3.1-8B-Instruct | 0.44 | Not recorded | 68fa7bb31e209131cea89ffa |
| DeepSeek-R1 | 0.29 | Not recorded | 68fa7bb31e209131cea89ffa |
| Claude 3.7 Sonnet | 0.28 | Not recorded | 68fa7bb31e209131cea89ffa |
| GPT-4o | 0.15 | Not recorded | 68fa7bb31e209131cea89ffa |
| Llama-3.1-8B-Instruct | 0.12 | Not recorded | 68fa7bb31e209131cea89ffa |
| Mistral-7B-Instruct-v0.3 | 0.11 | Not recorded | 68fa7bb31e209131cea89ffa |
| Qwen2.5-7B-Instruct | 0.08 | Not recorded | 68fa7bb31e209131cea89ffa |
| Llama-3.3-70B-Instruct | 0.08 | Not recorded | 68fa7bb31e209131cea89ffa |

### DeepSeek-R1

Persuasion Gain: 1.33. Result ID: 69afc7f96dbb7075b957bcb4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a91e209131cea6e3eb.md>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Claude 3.7 Sonnet

Persuasion Gain: 1.13. Result ID: 69afc7f96dbb7075b957bcb2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### GPT-4o

Persuasion Gain: 0.73. Result ID: 69afc7f96dbb7075b957bcb0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75681e209131cea67a98.md>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Mistral-7B-Instruct-v0.3

Persuasion Gain: 0.6. Result ID: 69afc7f96dbb7075b957bcaa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d61e209131cea4d909.md>)

[Implementation](<https://github.com/mistralai/mistral-src>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Qwen2.5-7B-Instruct

Persuasion Gain: 0.51. Result ID: 69afc7f96dbb7075b957bcac.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa767e1e209131cea6d6a7.md>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Llama-3.3-70B-Instruct

Persuasion Gain: 0.49. Result ID: 69afc7f96dbb7075b957bcae.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Llama-3.1-8B-Instruct

Persuasion Gain: 0.44. Result ID: 69afc7f96dbb7075b957bca8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Setting":"Dynamic","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### DeepSeek-R1

Persuasion Gain: 0.29. Result ID: 69afc7f96dbb7075b957bcb3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a91e209131cea6e3eb.md>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Claude 3.7 Sonnet

Persuasion Gain: 0.28. Result ID: 69afc7f96dbb7075b957bcb1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### GPT-4o

Persuasion Gain: 0.15. Result ID: 69afc7f96dbb7075b957bcaf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75681e209131cea67a98.md>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Llama-3.1-8B-Instruct

Persuasion Gain: 0.12. Result ID: 69afc7f96dbb7075b957bca7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Mistral-7B-Instruct-v0.3

Persuasion Gain: 0.11. Result ID: 69afc7f96dbb7075b957bca9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d61e209131cea4d909.md>)

[Implementation](<https://github.com/mistralai/mistral-src>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Qwen2.5-7B-Instruct

Persuasion Gain: 0.08. Result ID: 69afc7f96dbb7075b957bcab.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa767e1e209131cea6d6a7.md>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

### Llama-3.3-70B-Instruct

Persuasion Gain: 0.08. Result ID: 69afc7f96dbb7075b957bcad.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb31e209131cea89ffa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Setting":"Static","Receiver Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction.

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