Sampling dispreferred completions proportionally to the current reward model, as in contrastive divergence, improves preference-optimization performance and is framed as NLL estimation.
Anchored preference optimization and contrastive revisions: Addressing underspecification in alignment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Preference Optimization via Contrastive Divergence: Your Reward Model is Secretly an NLL Estimator
Sampling dispreferred completions proportionally to the current reward model, as in contrastive divergence, improves preference-optimization performance and is framed as NLL estimation.