PaperAudit-Bench - Section-Level Detection@1: leaderboard

Metric: F1 (%; harmonic mean of error coverage and finding precision on 1,209 corrupted sections of the ICML branch, the model prompted to report only its single most confident error; Qwen3-80B matching judge). Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.

Top models

#ModelScore
1Gemini 2.5 Pro50.6
2Claude Sonnet 4.548.3
3Qwen 3 235B A22B45.9
4DeepSeek V3.130.6
5Qwen 3 14B18.5
6Qwen 3 8B16.6

Interactive version: theaggregate.ai/benchmark?slug=paperaudit-bench-section-level-detection-1 · How It Works · Data refreshed daily, snapshot 2026-09-26.