AI RESEARCH
Targeted Tests for LLM Reasoning: An Audit-Constrained Protocol
arXiv CS.LG
•
ArXi:2605.11599v1 Announce Type: new Fixed reasoning benchmarks evaluate canonical prompts, but semantically valid changes in presentation can still change model behavior. Studies of prompt variation can reveal such failures, but without audit they can mix genuine model errors with invalid perturbations, extraction artifacts, and unmatched search procedures. We propose an audit-constrained protocol for targeted reasoning evaluation.