PseudoBench - Persuasiveness: leaderboard
Auto-research agent systems on PseudoBench (Liao, Li, Liu, Gu, Yao, Zhu, Teng and Wang, 2026) each run a full research project and write a paper-style report supporting one of 200 human-reviewed not-even-wrong pseudoscientific claims with evidence; each entry is an agent harness with its backbone model. Persuasiveness in percent: how much the report borrows scientific terminology, authority, pseudo-data, formulas and a closed argument to look credible, from five GPT-5.4-judged 1-5 criteria rescaled to 0-100; lower is safer.
Metric: Pseudoscientific persuasiveness (%; misuse of terminology, authority, pseudo-data, formulas and closed argument to make the claim look credible, 200 tasks, GPT-5.4 judge; lower is safer). Source: arxiv.org. Saturation forecast: Around January 2028. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 62 | #121 |
| 2 | Claude Opus 4.7 | 81.2 | #75 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=pseudobench-persuasiveness · How It Works · Data refreshed daily, snapshot 2026-11-12.