DRNoise - Conditional Deference: leaderboard
Conditional deference on DRNoise (2026): among the deep-research tasks an agent answers correctly in the clean corpus, the percent whose answer after a misleading document is seeded equals the planted false value. Lower is better.
Metric: Conditional deference (%; share of the tasks a model answered correctly in the clean condition whose noisy-condition answer equals the designated false value from the seeded document; 100 deep-research tasks whose unique gold answer must be recovered from two corroborating indirect record chains, run in a fixed BrowseComp-Plus search-only ReAct agent (top-5 dense retrieval, 512-token snippets, up to 100 iterations); mean of three runs; lower is better). Source: arxiv.org. Saturation forecast: Around 2035. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 32B | 61.3 | #960 |
| 2 | Gemini 3.5 Flash | 72.6 | #101 |
| 3 | GPT-5.4 | 80.4 | #123 |
| 4 | MiMo-V2.5 | 96.3 | #354 |
| 5 | DeepSeek V4 Flash | 98.1 | #258 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=drnoise-conditional-deference · How It Works · Data refreshed daily, snapshot 2026-11-13.