HM-Bench (VSR2): leaderboard
Metric: Average accuracy (%) over all tasks of HM-Bench's 19,337 hyperspectral multiple-choice questions (2,178 samples, 13 task categories) with full VSR2 input: the RGB, PCA and report views together with the authors' fixed cross-view fusion prompt; zero-shot, one option letter as output, temperature 0, at most 64 new tokens; higher is better. Source: arxiv.org. Saturation forecast: Around 2034. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 Mini | 44.28 |
| 2 | Claude Sonnet 4.6 | 44.01 |
| 3 | InternVL3-14B | 40.05 |
| 4 | Grok 4 | 39.5 |
| 5 | Qwen 2.5 VL 7B Instruct | 39.41 |
| 6 | InternVL2-8B | 39.12 |
Interactive version: theaggregate.ai/benchmark?slug=hm-bench-vsr2 · How It Works · Data refreshed daily, snapshot 2026-10-07.