Reward-weighted classifier-free guidance approximates Q-function policy improvement in autoregressive models, enabling test-time reward optimization and faster RL convergence via distillation.
Scores normalized so 0 = baseline model, 1 =r(y ∗)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Reward Weighted Classifier-Free Guidance as Policy Improvement in Autoregressive Models
Reward-weighted classifier-free guidance approximates Q-function policy improvement in autoregressive models, enabling test-time reward optimization and faster RL convergence via distillation.