EXPLORE-Bench - Attribute Score: leaderboard
Metric: Attribute-level score (0-5): a Qwen3-8B scorer rates the predicted attributes of each matched object against the annotation, averaged over matched objects and all 1,157 instances; the model sees an initial egocentric scene image and a sequence of atomic actions (11 to 694) from real Ego4D, Ego-Exo4D and self-recorded videos and describes the final scene; greedy decoding; higher is better. Source: arxiv.org. Saturation forecast: Around 2033. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 8B Instruct | 1.85 | #401 |
| 2 | Qwen 3 VL 8B (Thinking) | 1.84 | |
| 3 | Gemini 3 Flash | 1.8 | #93 |
| 4 | Gemini 3 Pro | 1.77 | #77 |
| 5 | GPT-5.2 Instant | 1.7 | #205 |
| 6 | Step3 VL 10B | 1.69 | #465 |
| 7 | Qwen 2.5 VL 7B Instruct | 1.48 | #643 |
| 8 | Qwen 2 VL 7B Instruct | 1.3 | #816 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=explore-bench-attribute-score · How It Works · Data refreshed daily, snapshot 2026-10-11.