SUGAR uses semantic entropy thresholds to dynamically switch between no retrieval, single-step retrieval, and multi-step retrieval, and reports accuracy gains on QA benchmarks.
Unsupervised dense information retrieval with con- trastive learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
REJECT 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
SUGAR: Leveraging Contextual Confidence for Smarter Retrieval
SUGAR uses semantic entropy thresholds to dynamically switch between no retrieval, single-step retrieval, and multi-step retrieval, and reports accuracy gains on QA benchmarks.