EgoCoT-Bench - State Retrospection: leaderboard
Metric: Answer accuracy (%, four-way multiple choice, strict exact match) of zero-shot MLLMs on egocentric manipulation videos of EgoCoT-Bench (351 clips, 3,172 human-verified questions built from spatio-temporal scene graphs), State Retrospection subtask (recall an earlier state of an object); higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 27B | 72.83 |
| 2 | Qwen 3 VL 235B A22B | 71.32 |
| 3 | Qwen 3.5 122B A10B | 70.72 |
| 4 | Qwen 3.5 397B A17B | 69.81 |
| 5 | GPT-5.2 | 67.55 |
| 6 | Qwen 3.5 Plus | 67.55 |
| 7 | GPT-5.1 | 66.04 |
Interactive version: theaggregate.ai/benchmark?slug=egocot-bench-state-retrospection · How It Works · Data refreshed daily, snapshot 2026-10-07.