Benchmark RadarAI BENCHMARK PROFILE
MusTBENCH
MusTBENCH evaluates temporal grounding in Large Audio-Language Models through five temporally grounded question-answering tasks, validated by music experts, to test alignment with audio regions.
- Released
- 2026-05-28
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Temporal grounding is critical for music understanding, where events are localized. MusTBENCH establishes this as a missing capability and offers a challenging benchmark for improvement.
Motivation
Recent Large Audio-Language Models (LALMs) have demonstrated promising abilities in understanding musical content.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.