Preprint
Aug 2026
TRACE: An Evidence-Grounded Benchmark for Safety Evaluation of Large Reasoning Models
TRACE is introduced, an evidence-grounded safety evaluation benchmark that covers the entire LRM inference pipeline: prompts, reasoning traces, and final responses, and reveals that safety judgment for reasoning traces is substantially more challenging than for prompts or final responses, and that current models struggle to accurately extract supporting evidence.
Zhenyu Wu, Siyu Chen, Changchun Yang et al.
· 0 citations