BOUND distills teacher-constructed corrective preferences at decision states where a locally plausible continuation would reinforce search drift, gaining up to 5.6 EM on Bamboogle and 4.8 accuracy on BrowseComp-Plus over Trajectory SFT.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
BOUND: Brief-Guided Corrective Preference Distillation at Search-Control Boundaries
BOUND distills teacher-constructed corrective preferences at decision states where a locally plausible continuation would reinforce search drift, gaining up to 5.6 EM on Bamboogle and 4.8 accuracy on BrowseComp-Plus over Trajectory SFT.