Benchmark
TRAPSBench
What is TRAPSBench?
TRAPSBench is a procedurally generated video benchmark with 1,404 physics pairs to test epistemic restraint in VLMs, using Penalized Epistemic Calibration…
- Released
- 2026-08-13
- Evaluates
- General AI, Multimodal
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
- Reported scores
- 0
Source provenance
- Original evidence https://arxiv.org/abs/2608.13167
TRAPSBench paper
No reported scores are on record for this benchmark yet.