SciHazard (Science-Specific Models) - Harm When Not Refused: leaderboard

Metric: DeHarm-Score over non-refused answers (0-5, lower is safer; the mean over the non-refused answers to 2,400 hazardous questions in 12 scientific disciplines of hazard level 1 to 5 x (executability + net-new risk, each 1 to 4)/8, executability from importance-weighted checklists and net-new risk from retrieval-checked claims, judged by Qwen3.5-35B-A3B; proprietary science-specific models and agents). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 3 models tracked.

Top models

#ModelScore
1InternAgent2.22
2S1-VL-32B2.68
3Intern-S1-Pro2.76

Interactive version: theaggregate.ai/benchmark?slug=scihazard-science-specific-models-harm-when-not-refused · How It Works · Data refreshed daily, snapshot 2026-09-29.