EgoIntrospect - Proactive Timing Judgment: leaderboard
Metric: Macro-F1 (%) over the interruptible and do-not-disturb classes of a binary judgment of whether a proactive assistant may interrupt the wearer during a clip (269 clips, labels from the wearers' own do-not-disturb annotations, interruptible the majority class), 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: Not forecast. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 8B Instruct | 60.69 |
| 2 | Kimi K2.5 | 59.78 |
| 3 | Qwen 3.5 397B A17B | 58.36 |
| 4 | Qwen3 Omni 30B A3B Instruct | 57.62 |
| 5 | Qwen 3 VL 8B (Thinking) | 57.57 |
| 6 | Gemini 3 Flash (Preview) | 57.41 |
| 7 | Gemini 3.1 Pro (Preview) | 57.13 |
| 8 | InternVL3-8B | 56.22 |
| 9 | Kimi K2.5 (Thinking) | 54.21 |
| 10 | Qwen 3.5 Omni Plus | 53.22 |
| 11 | GPT-4o | 29.55 |
| 12 | Qwen 2.5 VL 7B Instruct | 28.27 |
Interactive version: theaggregate.ai/benchmark?slug=egointrospect-proactive-timing-judgment · How It Works · Data refreshed daily, snapshot 2026-10-07.