EgoMemReason - Event Linking: leaderboard
EgoMemReason task (Wang et al., COLM 2026) that asks for the earlier event matching several constraints, such as place, activity type and time of day, among similar events in week-long egocentric video. Accuracy in percent on 100 multiple-choice questions; higher is better.
Metric: Accuracy (%; 100 multiple-choice questions). Source: arxiv.org. Saturation forecast: Around July 2028. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 44 | #159 |
| 2 | Gemini 3.1 Pro (Preview) | 33 | #91 |
| 3 | Molmo2-8B | 25 | #1488 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=egomemreason-event-linking · How It Works · Data refreshed daily, snapshot 2026-11-12.