A Practitioner's Guide to Multi-turn Agentic Reinforcement Learning

• URL: http://arxiv.org/abs/2510.01132v1
• Date: Wed, 01 Oct 2025 17:23:04 GMT
• Title: A Practitioner's Guide to Multi-turn Agentic Reinforcement Learning
• Authors: Ruiyi Wang, Prithviraj Ammanabrolu,
• Abstract summary: We study what actually works and what doesn't for training large language models as agents via multi-turn reinforcement learning.<n>We break down the design space into three inter-related pillars -- environment, reward, and policy.<n>We distill these findings into a training recipe that guides co-design across the three pillars.
• Score: 12.179148605060298
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We study what actually works and what doesn't for training large language models as agents via multi-turn reinforcement learning. Despite rapid progress, existing frameworks and definitions are fragmented, and there is no systematic formulation or analysis of which design choices matter across tasks. We address this gap by first breaking down the design space into three inter-related pillars -- environment, reward, and policy -- and empirically derive a recipe for training LLM agents in situated textual domains. In particular, we test TextWorld and ALFWorld, popular domains for testing situated embodied reasoning, as well as SWE-Gym for more software engineering style tasks. (i) For the environment, we analyze the impacts of task complexity in terms of sizes of the state and action spaces as well as optimal solution length, finding that even simple environments within a domain can provide signal on how well an agent can generalize to more complex tasks. (ii) For the reward, we ablate relative reward sparsity, observing that while dense turn-level rewards accelerate training, performance and stability is highly dependent on the choice of RL algorithm. (iii) And for the agent's policy, we explore the interplay between reward sparsity and biased (PPO, GRPO) and unbiased (RLOO) policy gradient methods in addition to showing how to find the optimal Supervised Fine-tuning (SFT) to RL training ratio given a fixed budget. We distill these findings into a training recipe that guides co-design across the three pillars, facilitating research and practical efforts in multi-turn agentic RL. Code: https://github.com/pearls-lab/meow-tea-taro

Related papers

• Sample-Efficient Neurosymbolic Deep Reinforcement Learning [49.60927398960061]
  We propose a neuro-symbolic Deep RL approach that integrates background symbolic knowledge to improve sample efficiency.<n>Online reasoning is performed to guide the training process through two mechanisms.<n>We show improved performance over a state-of-the-art reward machine baseline.
  arXiv  Detail & Related papers  (2026-01-06T09:28:53Z)
• Demystifying Reinforcement Learning in Agentic Reasoning [90.3737088727791]
  We conduct a comprehensive and systematic investigation to demystify reinforcement learning in agentic reasoning.<n>We highlight our key insights: (i) replacing stitched synthetic trajectories with real end-to-end tool-use trajectories yields a far stronger SFT.<n> Exploration-friendly techniques are crucial for agentic RL, such as clip higher, overlong reward shaping, and maintaining adequate policy entropy could improve the training efficiency.
  arXiv  Detail & Related papers  (2025-10-13T17:57:15Z)
• Agentic Reinforcement Learning with Implicit Step Rewards [92.26560379363492]
  Large language models (LLMs) are increasingly developed as autonomous agents using reinforcement learning (agentic RL)<n>We introduce implicit step rewards for agentic RL (iStar), a general credit-assignment strategy that integrates seamlessly with standard RL algorithms.<n>We evaluate our method on three challenging agent benchmarks, including WebShop and VisualSokoban, as well as open-ended social interactions with unverifiable rewards in SOTOPIA.
  arXiv  Detail & Related papers  (2025-09-23T16:15:42Z)
• AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learning [129.44038804430542]
  We introduce AgentGym-RL, a new framework to train LLM agents for multi-turn interactive decision-making through RL.<n>We propose ScalingInter-RL, a training approach designed for exploration-exploitation balance and stable RL optimization.<n>Our agents match or surpass commercial models on 27 tasks across diverse environments.
  arXiv  Detail & Related papers  (2025-09-10T16:46:11Z)
• Reinforcement Learning Foundations for Deep Research Systems: A Survey [31.57262766437479]
  This survey is the first dedicated to the RL foundations of deep research systems.<n>It systematizes work after DeepSeek-R1 along three axes: (i) data synthesis and curation; (ii) RL methods for agentic research covering stability, sample efficiency, long context handling, reward and credit design, multi-objective optimization, and multimodal integration; and (iii) agentic RL training systems and frameworks.
  arXiv  Detail & Related papers  (2025-09-08T14:27:23Z)
• Agentic Reinforced Policy Optimization [66.96989268893932]
  Large-scale reinforcement learning with verifiable rewards (RLVR) has demonstrated its effectiveness in harnessing the potential of large language models (LLMs) for single-turn reasoning tasks.<n>Current RL algorithms inadequately balance the models' intrinsic long-horizon reasoning capabilities and their proficiency in multi-turn tool interactions.<n>We propose Agentic Reinforced Policy Optimization (ARPO), a novel agentic RL algorithm tailored for training multi-turn LLM-based agents.
  arXiv  Detail & Related papers  (2025-07-26T07:53:11Z)
• Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Reward Design [35.544075583073685]
  We present the first systematic study of textitturn-level reward design for multi-turn RL algorithms and agent applications.<n>We conduct case studies on multi-turn reasoning-augmented search agents, where we carefully design two types of turn-level rewards: verifiable and LLM-as-judge.<n>Our experiments on multi-turn search tasks demonstrate that incorporating well-designed turn-level rewards enables RL algorithms to significantly outperform baseline methods with trajectory-level rewards.
  arXiv  Detail & Related papers  (2025-05-17T04:09:46Z)
• RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learning [125.96848846966087]
  Training large language models (LLMs) as interactive agents presents unique challenges.<n>While reinforcement learning has enabled progress in static tasks, multi-turn agent RL training remains underexplored.<n>We propose StarPO, a general framework for trajectory-level agent RL, and introduce RAGEN, a modular system for training and evaluating LLM agents.
  arXiv  Detail & Related papers  (2025-04-24T17:57:08Z)
• Guiding Multi-agent Multi-task Reinforcement Learning by a Hierarchical Framework with Logical Reward Shaping [16.5526277899717]
  This study aims to design a multi-agent cooperative algorithm with logic reward shaping. Experiments have been conducted on various types of tasks in the Minecraft-like environment.
  arXiv  Detail & Related papers  (2024-11-02T09:03:23Z)
• Exploring RL-based LLM Training for Formal Language Tasks with Programmed Rewards [49.7719149179179]
  This paper investigates the feasibility of using PPO for reinforcement learning (RL) from explicitly programmed reward signals. We focus on tasks expressed through formal languages, such as programming, where explicit reward functions can be programmed to automatically assess quality of generated outputs. Our results show that pure RL-based training for the two formal language tasks is challenging, with success being limited even for the simple arithmetic task.
  arXiv  Detail & Related papers  (2024-10-22T15:59:58Z)
• Spatial Reasoning and Planning for Deep Embodied Agents [2.7195102129095003]
  This thesis explores the development of data-driven techniques for spatial reasoning and planning tasks. It focuses on enhancing learning efficiency, interpretability, and transferability across novel scenarios.
  arXiv  Detail & Related papers  (2024-09-28T23:05:56Z)
• PEAR: Primitive Enabled Adaptive Relabeling for Boosting Hierarchical Reinforcement Learning [25.84621883831624]
  Hierarchical reinforcement learning (HRL) has the potential to solve complex long horizon tasks using temporal abstraction and increased exploration.<n>We present primitive enabled adaptive relabeling (PEAR)<n>We first perform adaptive relabeling on a few expert demonstrations to generate efficient subgoal supervision.<n>We then jointly optimize HRL agents by employing reinforcement learning (RL) and imitation learning (IL)
  arXiv  Detail & Related papers  (2023-06-10T09:41:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.