Counterfactual-World (CF-World) Benchmark
Explore score history ↗A three-level benchmark evaluating whether text-to-image models can follow explicit and implicit counterfactual rules rather than defaulting to real-world visual priors.
- Released
- 2026-06-23
- Institutions
- Shanghai Jiao Tong University · Shanghai Artificial Intelligence Laboratory · Chinese University of Hong Kong
- Evaluates
- General AI, Reasoning, Vision & 3D
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- ai-reviewed
Counterfactual-World (CF-World) paper and code
No reported scores are on record for this benchmark yet.