SurveyReview - Comprehensiveness: leaderboard
Mean squared error of predicted comprehensiveness review scores in SurveyReview (2026), from 0 to 16; lower is better.
Metric: Mean squared error of predicted comprehensiveness scores (0 to 16, lower is better; zero-shot LLM-as-a-judge with the unified dimension-specific rubric prompt on the full survey text, against scores annotated from real peer-review reports of the test-split surveys (MOPRD, F1000Research, OpenReview) on the {-2, -1, +1, +2} scale). Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 2.04 | #172 |
| 2 | Claude Opus 4.5 (Thinking) | 2.66 | #152 (Claude Opus 4.5) |
| 3 | Qwen 3 32B | 3.22 | #960 |
| 4 | GLM-4.7 | 3.66 | #365 |
| 5 | Gemini 3 Pro (Preview) | 3.91 | #111 |
| 6 | DeepSeek V3.2 | 4.59 | #403 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=surveyreview-comprehensiveness · How It Works · Data refreshed daily, snapshot 2026-11-12.