SciHazard Benchmark

What is SciHazard?

SciHazard evaluates LLMs on scientific safety risks across 12 disciplines with 2,400 hazardous and 600 oversafety questions grounded in regulated entities. The…

Released

2026-07-21

Evaluates

General AI, Safety, Safety & Trustworthiness

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.18665

SciHazard paper

• Paper https://arxiv.org/abs/2607.18665
• Website https://anonymous.4open.science/r/DeharmScore-7B55

Which sources cite SciHazard?

• SciHazard

Related General AI, Safety, Safety & Trustworthiness benchmarks

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