Scientia AI●

a bilingual, research-first reading log for AI

auth

← back to the log

arXiv cs.LG

Predictable Failure in Multi-Hop Retrieval: Score-Distributional Confidence Scoring and Abstention

다중 홉 검색에서 예측 가능한 실패: 점수 분포 신뢰도 채점과 기피

2026.09.18 · general-ai

Read the original article ↗

No write-up yet — this entry is still just the auto-fetched hook. Aaron's own summary (200–500 words) will replace this note once he's written it.