슬라이드 3장11:59ENMBMatthew Berman· 2026년 3월 11일How Claude Identified a Benchmark and Retrieved Its AnswersMatthew Berman · Forward FutureMatthew Berman examines Anthropic's report on BrowseComp contamination, evaluation-aware behavior, and the difficulty of keeping web-based tests reliable.평가(Evals)에이전트 엔지니어링Anthropic+2