Temporally Coherent Reward Modeling adds Monte Carlo and TD regularizers to Bradley-Terry training so reward model outputs at every token become conditional expectations of the final reward, improving token-level interpretability, process supervision from outcome data, and PPO efficiency.
Sincey0..k is more informative thany0..k−1, we should expect to see the mean squared prediction error(r(x, y0..k)−r(x, y)) 2 decrease askincreases
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 Models Are Secretly Value Functions: Temporally Coherent Reward Modeling
Temporally Coherent Reward Modeling adds Monte Carlo and TD regularizers to Bradley-Terry training so reward model outputs at every token become conditional expectations of the final reward, improving token-level interpretability, process supervision from outcome data, and PPO efficiency.