Appendix for " Preference-grounded Token-level Guidance for 657 Language Model Fine-tuning " 658 Table of Contents
–Neural Information Processing Systems
F.3 Sparse Reward with KL Penalty . . . . . . . . . . . . . . . . . . . . . . . . . . .
Neural Information Processing Systems
Feb-11-2026, 14:07:34 GMT