CausalVQA Benchmark
Explore score history ↗CausalVQA tests causal reasoning in videos across five question types, and state-of-the-art multimodal models still trail human performance.
- Publisher
- FAIR at Meta (published the hub card)
- Released
- 2025-06-11
- Modality
- multimodal
- Evaluates
- 多模态, Multimodal, VQA, 多模态模型, VLM, 跨模态推理, Cross-modal Reasoning
- Openness
- unknown
- Importer
- OpenCompass Hub
Source provenance
CausalVQA paper and code
No reported scores are on record for this benchmark yet.