EgoCoT-Bench - Hand-Object Temporal Retrospection: leaderboard
EgoCoT-Bench (2026) Hand-Object Temporal Retrospection subtask: from an egocentric manipulation video, order earlier hand-object interactions in time. 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), Hand-Object Temporal Retrospection subtask (order earlier hand-object interactions in time); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 19 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | InternVL3.5-14B | 75 | #1519 |
| 2 | InternVL3.5-1B | 75 | #6140 |
| 3 | InternVL3.5-2B | 71.88 | #3313 |
| 4 | InternVL3.5-8B | 68.75 | #1618 |
| 5 | Qwen 3 VL 32B | 68.75 | #624 |
| 6 | InternVL3.5-4B | 62.5 | #1946 |
| 7 | Qwen 3 VL 30B A3B | 62.5 | #791 |
| 8 | GPT-5.2 | 59.38 | #190 |
| 9 | Qwen 3.5 27B | 59.38 | #380 |
| 10 | Qwen 3 VL 8B | 59.38 | #927 |
| 11 | Qwen 3.5 397B A17B | 56.25 | #270 |
| 12 | Qwen 3.5 122B A10B | 56.25 | #350 |
| 13 | GPT-5.1 | 56.25 | #248 |
| 14 | Qwen 3 VL 235B A22B | 56.25 | #486 |
| 15 | Qwen 3 VL Plus | 56.25 | #514 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=egocot-bench-hand-object-temporal-retrospection · How It Works · Data refreshed daily, snapshot 2026-11-22.