MiraBench

What is MiraBench?

MiraBench evaluates action-conditioned reliability in robotic world models through three levels: Physics Adherence, Action-Following Fidelity, and Optimism…

Released

2026-05-28

Evaluates

Language & Knowledge, Robotics & Embodied AI, cs.AI

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2605.29360

MiraBench paper

• Paper https://arxiv.org/abs/2605.29360

Which sources cite MiraBench?

• MiraBench

Related Language & Knowledge, Robotics & Embodied AI, cs.AI benchmarks

• 360CityArena
• 4DSynth-Nav
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• AARRI-Bench
• ABC-Bench
• ActiveFly-Bench
• ActiveSciBench