EgoIntrospect - Memory Recall Prediction: leaderboard
Metric: Accuracy (%) on 200 four-choice questions asking which detail of a 3-minute clip the wearer later recalled as vivid (distractors are other elements of the same clip written by Gemini and curated by the authors), fixation-overlaid egocentric clips from 15 held-out participants with the participant profile; native audio-video models hear the clip audio, the other models read its transcript; temperature 0 (0.6 for InternVL3.5 thinking mode); higher is better. Source: arxiv.org. Saturation forecast: Around 2032. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen3 Omni 30B A3B Instruct | 55.5 |
| 2 | Qwen 3.5 Omni Plus | 52 |
| 3 | GPT-4o | 51.5 |
| 4 | Gemini 3.1 Pro (Preview) | 50 |
| 5 | Qwen 3.5 397B A17B | 49 |
| 6 | Kimi K2.5 (Thinking) | 48.5 |
| 7 | Kimi K2.5 | 47.5 |
| 8 | Qwen 3 VL 8B Instruct | 47.5 |
| 9 | Gemini 3 Flash (Preview) | 47 |
| 10 | Qwen 3 VL 8B (Thinking) | 43 |
| 11 | InternVL3-8B | 40.5 |
| 12 | Qwen 2.5 VL 7B Instruct | 38 |
Interactive version: theaggregate.ai/benchmark?slug=egointrospect-memory-recall-prediction · How It Works · Data refreshed daily, snapshot 2026-10-07.