EgoCoT-Bench - State Retrospection: leaderboard
EgoCoT-Bench (2026) State Retrospection subtask: from an egocentric manipulation video, recall an earlier state of an object. Four-way multiple choice accuracy in percent; higher is better.
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. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 27B | 72.83 | #375 |
| 2 | Qwen 3 VL 235B A22B | 71.32 | #485 |
| 3 | Qwen 3.5 122B A10B | 70.72 | #346 |
| 4 | Qwen 3.5 397B A17B | 69.81 | #264 |
| 5 | Qwen 3 VL Plus | 69.7 | #512 |
| 6 | GPT-5.2 | 67.55 | #185 |
| 7 | Qwen 3.5 Plus | 67.55 | #242 |
| 8 | GPT-5.1 | 66.04 | #240 |
| 9 | Qwen 3 VL 30B A3B | 65.15 | #784 |
| 10 | Qwen 3 VL 32B | 64.53 | #622 |
| 11 | Qwen 3 VL 8B | 64.02 | #917 |
| 12 | LLaVA-OneVision-1.5-4B | 61.51 | #1855 |
| 13 | InternVL3.5-2B | 58.11 | #3284 |
| 14 | LLaVA-OneVision-1.5-8B | 57.36 | #1864 |
| 15 | InternVL3.5-14B | 56.98 | #1504 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=egocot-bench-state-retrospection · How It Works · Data refreshed daily, snapshot 2026-11-21.