BASS - Pairwise-Gene Detection: leaderboard
Metric: Exact-match accuracy (%) on the 147 pairwise-gene detection questions: pick the pair of most prominent attributes among eight, drawn from usually anticorrelated attributes (multiple-choice, majority vote over four runs); BASS music understanding benchmark, audio language models get a text prompt (one of ten paraphrases) and a full commercial music recording from BASS's songs; raw score before chance normalization; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 10.88 | #145 |
| 2 | Voxtral-Small-24B-2507 | 8.84 | #565 |
| 3 | Phi-4 Multimodal Instruct | 1.36 | #896 |
| 4 | SALMONN-7B | 1.36 | #1274 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=bass-pairwise-gene-detection · How It Works · Data refreshed daily, snapshot 2026-10-11.