Benchmark contains 392 videos and 1,027 question-answer pairs for audio-visual reasoning
Researchers created a benchmark containing 392 videos and 1,027 human-annotated question-answer pairs to evaluate audio-visual reasoning in multimodal language models.
The benchmark requires models to return both an answer and supporting evidence intervals to evaluate the complete reasoning chain from audio and visual information.
Sources behind this report
- Hugging Face
Research · abstract · CC0 1.0
Source published
Read by the newsroom
Report history
No earlier revision dates are retained here. This does not establish that the report has never changed; read any dated updates and corrections.
This brief was made by AI and checked against its cited public sources. How the newsroom works