PhysDox: Benchmarking LLMs on Physical Feasibility Auditing of Physiological Sensing Protocols
For researchers using LLMs in experimental design, this benchmark reveals that protocol auditing requires calibrated feasibility reasoning, not factual recall.
LLMs struggle to audit physical feasibility of biomedical protocols, achieving only 53.0 macro-F1 on severity detection, with implicit constraints missed twice as often as explicit ones.
Large language models (LLMs) increasingly assist in experimental design, yet fluent protocols often remain physically infeasible. We introduce PhysDox, a physical feasibility auditing benchmark for biomedical protocols comprising a 683-sample expert-curated Gold set and a 5,000-sample Silver set across six sensing domains. We formulate the task as a two-stage evaluation: severity detection classifying protocols as valid, minor, or fatal, followed by the constraint-level diagnosis of fatal violations. Evaluating 6 LLMs across 4 inference strategies yields a peak Stage-1 macro-F1 of only 53.0. Moreover, strong oracle diagnosis collapses during end-to-end evaluation due to correlated cascade errors. Error analysis reveals scaffold bias, where models conflate procedural completeness with physical validity. Consequently, implicit constraints exhibit a 2 times higher miss rate than explicit hardware violations, supported by strong statistical correlation at $ρ{=}0.81$ and $p{<}0.01$. Trace analysis of false negatives exposes a 54%--46% split between attention and judgment failures, ultimately demonstrating that protocol auditing demands calibrated feasibility reasoning rather than factual recall or longer rationales.