{"type":"link","version":"1.0","title":"An LLM judge's asymmetric false-positive rate biases retrieval-success estimates upward, and more judge-labeled samples do not fix it","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/risk-llm-judges-overestimate-retrieval-success","thumbnail_url":"https://ai-arch.pages.dev/og/risk-llm-judges-overestimate-retrieval-success.png","thumbnail_width":1200,"thumbnail_height":630}