AI scientists can pass the test while their reasoning breaks
Two AI scientists can reach the same answer while taking radically different paths, and output-only benchmarks will miss it. OpenDiscoveryTrace publishes 558 full trajectories across 124 scientific tasks, logging tool calls, errors, revisions, and confidence. Its pilot found frontier models with similar success rates of 84% to 89%, yet Claude Opus 4.6 made 30 times more errors than GPT-5.4, exposing why process-level audits matter for scientific agents.