{"type":"link","version":"1.0","title":"Benchmarks that grade only agent intent miss systematic execution failures and overestimate task completion","author_name":"AI Archs","author_url":"https://ai-arch.pages.dev","provider_name":"AI Archs","provider_url":"https://ai-arch.pages.dev","url":"https://ai-arch.pages.dev/n/benchmark-evals-must-verify-execution-state-not-just-intent","thumbnail_url":"https://ai-arch.pages.dev/og/benchmark-evals-must-verify-execution-state-not-just-intent.png","thumbnail_width":1200,"thumbnail_height":630}