Distract-Bench

What is Distract-Bench?

Distract-Bench evaluates robustness of vision-language models to semantic visual distractions, which are meaningful but task-irrelevant cues that preserve the…

Released

2026-06-08

Evaluates

General AI, Multimodal, Reasoning, Robustness, Safety & Trustworthiness

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2606.08894

Distract-Bench paper and code

• Paper https://arxiv.org/abs/2606.08894
• Code https://github.com/Yizheng-Sun/Distract-Bench

Which sources cite Distract-Bench?

• Distract-Bench

Related General AI, Multimodal, Reasoning, Robustness, Safety & Trustworthiness benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A-Bench
• A-OKVQA
• A3S-Bench
• AARRI-Bench
• ActBench