ARLArena: A Unified Framework for Stable Agentic Reinforcement Learning
This work addresses the instability issue in ARL, which limits scalability and exploration, providing a unified perspective and practical guidance for building stable LLM-based agent training pipelines.
The paper tackles the instability problem in agentic reinforcement learning (ARL) by proposing ARLArena, a framework for stable training and analysis, and SAMPO, a method that achieves consistently stable training and strong performance across diverse tasks.
Agentic reinforcement learning (ARL) has rapidly gained attention as a promising paradigm for training agents to solve complex, multi-step interactive tasks. Despite encouraging early results, ARL remains highly unstable, often leading to training collapse. This instability limits scalability to larger environments and longer interaction horizons, and constrains systematic exploration of algorithmic design choices. In this paper, we first propose ARLArena, a stable training recipe and systematic analysis framework that examines training stability in a controlled and reproducible setting. ARLArena first constructs a clean and standardized testbed. Then, we decompose policy gradient into four core design dimensions and assess the performance and stability of each dimension. Through this fine-grained analysis, we distill a unified perspective on ARL and propose SAMPO, a stable agentic policy optimization method designed to mitigate the dominant sources of instability in ARL. Empirically, SAMPO achieves consistently stable training and strong performance across diverse agentic tasks. Overall, this study provides a unifying policy gradient perspective for ARL and offers practical guidance for building stable and reproducible LLM-based agent training pipelines.