EgoSafetyBench - Sequential Detection: leaderboard
Metric: Caught rate (%; share of unsafe situational videos whose first alarm falls within one half-second chunk after the annotated hazard onset, chunks streamed in order; ten VLM safety guards judging egocentric robot videos split into ten half-second chunks (ten frames each at 1280x720), each chunk judged alone under the current-chunk protocol with a single-word verdict and greedy decoding; higher is better). Source: arxiv.org. Saturation forecast: Around March 2028. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 57.4 |
| 2 | Gemini 3.5 Flash | 54.2 |
| 3 | Gemini 3.1 Flash Lite | 52.3 |
| 4 | Qwen 3 VL 4B Instruct | 47.7 |
| 5 | Qwen 3.5 4B | 45.6 |
| 6 | Gemma 3 4B (IT) | 43.5 |
| 7 | InternVL3.5-8B | 39.2 |
| 8 | Qwen 3.5 0.8B | 15.8 |
Interactive version: theaggregate.ai/benchmark?slug=egosafetybench-sequential-detection · How It Works · Data refreshed daily, snapshot 2026-09-29.