SPIKE-Bench

What is SPIKE-Bench?

SPIKE-Bench evaluates LLM biosecurity risks using 631 curated toxin-design prompts and a three-stage funnel (compliance, plausibility, predicted toxicity)…

Released

2026-08-03

Evaluates

Biology & Drug Discovery, Safety, Safety & Trustworthiness

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2608.02684

SPIKE-Bench paper and code

• Paper https://arxiv.org/abs/2608.02684
• Code https://github.com/PKU-Alignment/SPIKE-Bench

Which sources cite SPIKE-Bench?

• SPIKE-Bench

Related Biology & Drug Discovery, Safety, Safety & Trustworthiness benchmarks

• A Large-Scale Dataset and Benchmark
• ABC-Bench
• ActBench
• Agent Planning Benchmark
• AgentHarm
• AgentHazard
• AgentS4D
• AICompanionBench