Mr Dre - Format Feedback Incorporation Rate: leaderboard
Multi-turn report revision by deep research agents on Mr Dre (Chen et al., 2026): after a first report on an expert-written research question from ResearchRubrics, RigorousBench or ResearcherBench, the agent receives simulated user feedback asking for a change in presentation and revises the report. Incorporation rate in percent is the share of revisions that a GPT-4.1-mini judge finds follow the formatting request, averaged over the three question sets; higher is better.
Metric: Incorporation rate (%; turn-2 revision after format feedback: share of revisions a judge finds follow the formatting request; averaged over ResearchRubrics, RigorousBench and ResearcherBench questions; GPT-4.1-mini judge). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | OpenAI o4-mini Deep Research | 98.5 | |
| 2 | dr-tulu | 94.9 | |
| 3 | Tongyi-DeepResearch-30B-A3B | 94.3 | |
| 4 | LangChain Open Deep Research (GPT-4.1-mini) | 91.1 | |
| 5 | Sonar Deep Research | 85.5 |
Interactive version: theaggregate.ai/benchmark?slug=mr-dre-format-feedback-incorporation-rate · How It Works · Data refreshed daily, snapshot 2026-11-12.