FailBench

What is FailBench?

A robot failure detection benchmark with 2,197 manipulation attempts from real and simulated sources. Benchmark Radar tracks FailBench, cited by 1 source…

Released

2026-09-03

Evaluates

General AI, Robot manipulation, Robotics, Robotics & Embodied AI

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2609.03611

FailBench paper

• Paper https://arxiv.org/abs/2609.03611

Which sources cite FailBench?

• FailBench

Related General AI, Robot manipulation, Robotics, Robotics & Embodied AI benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 360CityArena
• 3DCodeBench
• 4DSynth-Nav
• 5-Dialects-BN
• A3S-Bench
• AARRI-Bench
• ABC-Bench