Benchmark Radar™
Star —

TRAPSBench

Explore score history ↗

TRAPSBench is a procedurally generated video benchmark with 1,404 physics pairs to test epistemic restraint in VLMs, using Penalized Epistemic Calibration Score (PECS).

Released
2026-08-13
Institutions
Meta (United Kingdom)
Evaluates
General AI, Multimodal
Openness
unknown
Importer
Claire Radar
Review status
unreviewed

TRAPSBench paper

No reported scores are on record for this benchmark yet.

Which sources cite TRAPSBench?

Related General AI, Multimodal benchmarks