Training-Free Binary Projection Filtering for Dense Retrieval: An Empirical Study of Candidate Reduction, Ranking Stability, and Failure Risk

Dense retrieval pipelines often rely on large candidate pools before reranking, making candidate generation and downstream scoring a practical bottleneck. This paper studies training-free binary projection filtering as a lightweight pre-filter for reducing the candidate set before dense reranking. R...

पूर्ण विवरण

ग्रंथसूची विवरण

मुख्य लेखकों:	Tip-aroon Kiawkaew, Thanaruk Theeramunkong
स्वरूप:	Artigo
भाषा:	Inglês
प्रकाशित:	MDPI AG 2026-05-01
श्रृंखला:	Information
विषय:	dense retrieval candidate reduction binary projection filtering dense reranking catastrophic failure retrieval robustness
ऑनलाइन पहुंच:	https://www.mdpi.com/2078-2489/17/5/514
टैग:	टैग जोड़ें कोई टैग नहीं, इस रिकॉर्ड को टैग करने वाले पहले व्यक्ति बनें!