analiza aiqbench
This paper presents DRIFT, a novel method for offline-to-online reinforcement learning that adapts continuous flow matching techniques to discrete action spaces using advantage-weighted losses and path-space regularization. By employing candidate-set approximations, the approach efficiently handles large discrete action spaces while preserving knowledge from static datasets during online fine-tuning. The results demonstrate superior performance over both online-only baselines and pretrained language model counterparts in macro-action MinAtar environments.
Discrete Flow Matching for Offline-to-Online Reinforcement Learning
arXiv:2605.12379v2 Announce Type: replace-cross Abstract: Many reinforcement learning (RL) tasks have discrete action spaces, but most generative policy methods based on diffusion and flow matching are designed for continuous control. Meanwhile, generative policies usually rely heavily on offline datasets and offline-to online RL is itself challenging, as the policy must improve from new interaction without losing useful behavior learned from static data. To address those challenges, we introduce DRIFT, an online fine-tuning method that updates an offline pretrained continuous time Markov chain (CTMC) policy with an advantage-weighted discrete flow matching loss. To preserve useful pretrained knowledge, we add a path-space penalty that regularizes the full CTMC trajectory distribution, rather than only the final action distribution. For large discrete action spaces, we introduce a candidate-set approximation that updates the actor over a small subset of actions sampled from reference-policy rollouts and uniform exploration. Our theoretical analysis shows that focusing on a small set of high probability actions can reduce computation while keeping the approximation error small, making our method practical for large action spaces. We evaluate on macro-action MinAtar games where offline-pretrained DRIFT outperforms its online-only counterpart across all five games and achieves the highest mean return on three and on oracle free Jericho, DRIFT achieves a mean normalized score of 7.3\% with a lightweight GRU encoder, outperforming the evaluated pretrained language-model baselines. Ablations demonstrate the importance of appropriately weighted path-space regularization and reference-guided candidate construction.
Abstract:Many reinforcement learning (RL) tasks have discrete action spaces, but most generative policy methods based on diffusion and flow matching are designed for continuous control. Meanwhile, generative policies usually rely heavily on offline datasets and offline-to online RL is itself challenging, as the policy must improve from new interaction without losing useful behavior learned from static data. To address those challenges, we introduce DRIFT, an online fine-tuning method that updates an offline pretrained continuous time Markov chain (CTMC) policy with an advantage-weighted discrete flow matching loss. To preserve useful pretrained knowledge, we add a path-space penalty that regularizes the full CTMC trajectory distribution, rather than only the final action distribution. For large discrete action spaces, we introduce a candidate-set approximation that updates the actor over a small subset of actions sampled from reference-policy rollouts and uniform exploration. Our theoretical analysis shows that focusing on a small set of high probability actions can reduce computation while keeping the approximation error small, making our method practical for large action spaces. We evaluate on macro-action MinAtar games where offline-pretrained DRIFT outperforms its online-only counterpart across all five games and achieves the highest mean return on three and on oracle free Jericho, DRIFT achieves a mean normalized score of 7.3\% with a lightweight GRU encoder, outperforming the evaluated pretrained language-model baselines. Ablations demonstrate the importance of appropriately weighted path-space regularization and reference-guided candidate construction.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2605.12379 [cs.LG]
(or arXiv:2605.12379v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2605.12379 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 325 cuvinte
Subiecte conexe
A Survey of Flow Matching in Reinforcement Learning
arXiv · 15 oct. 2026
Test-time Multi-agent Coordination by Decomposed Value Gradient Flow
arXiv · 5 oct. 2026
Flow Matching for Offline Reinforcement Learning with Discrete Actions
arXiv · 14 oct. 2026
Sidekick's continual learning loop
Shopify Engineering Blog · 5 aug. 2026