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

#ModelScore
1Qwen3 Omni 30B A3B Instruct55.5
2Qwen 3.5 Omni Plus52
3GPT-4o51.5
4Gemini 3.1 Pro (Preview)50
5Qwen 3.5 397B A17B49
6Kimi K2.5 (Thinking)48.5
7Kimi K2.547.5
8Qwen 3 VL 8B Instruct47.5
9Gemini 3 Flash (Preview)47
10Qwen 3 VL 8B (Thinking)43
11InternVL3-8B40.5
12Qwen 2.5 VL 7B Instruct38

Interactive version: theaggregate.ai/benchmark?slug=egointrospect-memory-recall-prediction · How It Works · Data refreshed daily, snapshot 2026-10-07.