paper-with-me

홈 › Papers

MDP Geometry, Normalization and Reward Balancing Solvers

2024-07-09 · Arsenii Mustafin, Aleksei Pakharev, Alex Olshevsky, Ioannis Ch. Paschalidis

We present a new geometric interpretation of Markov Decision Processes (MDPs) with a natural normalization procedure that allows us to adjust the value function at each state without altering the advantage of any action with respect to any policy. This advantage-preserving transformation of the MDP motivates a class of algorithms which we call Reward Balancing, which solve MDPs by iterating through these transformations, until an approximately optimal policy can be trivially found. We provide a convergence analysis of several algorithms in this class, in particular showing that for MDPs for unknown transition probabilities we can improve upon state-of-the-art sample complexity results.

📄 PDF Abstract BibTeX arXiv:2407.06712

Code (0)

등록된 구현이 없습니다.

Tasks

Decision MakingSequential Decision Making

Similar Papers 제목 키워드 기반

HPO: Hysteretic Policy Optimization for Stable and Efficient Training under Sparse-Reward Regime

2026-05-28 · Mohamed Sana, Nicola Piovesan, Antonio De Domenico, Fadhel Ayed 외 arxiv

We investigate a narrow but common failure mode of GRPO-style reinforcement learning in the context of sparse verifiable rewards: early updates contain more responses with negative advantages than those with positive adv…

Reinforcement Learning

MuonEq: Balancing Before Orthogonalization with Lightweight Equilibration

2026-03-30 · Da Chang, Qiankun Shi, Lvgang Zhang, Yu Li 외 arxiv

Orthogonalized-update optimizers such as Muon improve training of matrix-valued parameters, but existing extensions typically either rescale updates after orthogonalization or use heavier whitening-based preconditioners …

MAdam: Metric-Aware Multi-Objective Adam

2026-06-02 · Fengbei Liu, Rachit Saluja, Sunwoo Kwak, Ruibo Wang 외 arxiv

Multi-objective optimization (MOO) underlies many machine learning problems, yet MOO solvers across the loss-balancing, gradient-balancing, and Pareto-based families almost universally hand their reconciled directions to…

Multi-Task Learning

SALT: When More Rollouts Don't Help in Group-Based Policy Optimization and How to Make Them Matter

2026-06-04 · Powei Chang, Jinpeng Zhang, Chaoqun Sun, MiniWell Tsao 외 arxiv

Reinforcement learning with verifiable rewards (RLVR) often adopts GRPO-style group-relative updates, sampling multiple rollouts per prompt to construct normalized learning signals. However, merely increasing the number …

Reinforcement Learning

In-Dataset Trajectory Return Regularization for Offline Preference-based Reinforcement Learning

2024-12-12 · Songjun Tu, Jingbo Sun, Qichao Zhang, Yaocheng Zhang 외

Offline preference-based reinforcement learning (PbRL) typically operates in two phases: first, use human preferences to learn a reward model and annotate rewards for a reward-free offline dataset; second, learn a policy…

Offline RL