BeliefShift (RAG) - Evidence Sensitivity: leaderboard
Metric: Evidence Sensitivity Index times 100 (-100 to 100): the rate of meaningful belief updates in evidence sessions minus the rate in drift-pressure sessions without evidence, over BeliefShift trajectories, with the top-k earlier sessions retrieved by a dense index; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet | 63 | #337 |
| 2 | Gemini 1.5 Pro | 58 | #442 |
| 3 | GPT-4o | 42 | #333 |
| 4 | Mistral Large | 31 | #693 |
| 5 | Gemini 1.5 Flash | 24 | #619 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=beliefshift-rag-evidence-sensitivity · How It Works · Data refreshed daily, snapshot 2026-10-11.