Rectifying Belief Space via Unlearning to Harness LLMs' Reasoning.

Ayana Niwa, Masahiro Kaneko, Kentaro Inui

Browse the full ACL paper archive.