The cumulative token IS ratio gives unbiased prefix correction and lower variance than full-sequence ratios for token-level gradients in LLM policy optimization, enabling CTPO to outperform GRPO and GSPO baselines on mathematical reasoning tasks.
Improving LLM general preference alignment via optimistic online mirror descent
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
unclear 1representative citing papers
Introduces HRC model for game-theoretic decomposition of preferences into orthogonal transitive and cyclic components, paired with DSPPO for dynamic Nash-seeking alignment, reporting gains over BT and GPM baselines on RewardBench and downstream LLM evaluations.
Diff.-NPO frames diffusion alignment as a self-play game reaching Nash equilibrium and reports better text-to-image results than prior DPO-style methods.
An expository book that systematically presents RLHF methods, from reward modeling to direct alignment algorithms, aimed at readers with quantitative backgrounds.
citing papers explorer
-
Rethinking Importance Sampling in LLM Policy Optimization: A Cumulative Token Perspective
The cumulative token IS ratio gives unbiased prefix correction and lower variance than full-sequence ratios for token-level gradients in LLM policy optimization, enabling CTPO to outperform GRPO and GSPO baselines on mathematical reasoning tasks.
-
Transitivity Meets Cyclicity: Explicit Preference Decomposition for Dynamic Large Language Model Alignment
Introduces HRC model for game-theoretic decomposition of preferences into orthogonal transitive and cyclic components, paired with DSPPO for dynamic Nash-seeking alignment, reporting gains over BT and GPM baselines on RewardBench and downstream LLM evaluations.
-
Towards General Preference Alignment: Diffusion Models at Nash Equilibrium
Diff.-NPO frames diffusion alignment as a self-play game reaching Nash equilibrium and reports better text-to-image results than prior DPO-style methods.
-
Reinforcement Learning from Human Feedback
An expository book that systematically presents RLHF methods, from reward modeling to direct alignment algorithms, aimed at readers with quantitative backgrounds.
- Multiplayer Nash Preference Optimization