TensorX
返回文献探索

Paper · arXiv 2608.17528

Agent Lightning v1.0: Towards Harnessed Agentic RL

Zhiyuan He, Siwei Zhang, Zhiwen Zhou, Yuqing Yang, Yu Kang, Yuge Zhang, Luna K. Qiu, Tin Yan Tsui, Jiahang Xu, Chong Luo

35 upvotesAugust 18, 2026arXiv 预印本
AI 摘要

Agent Lightning v1.0 enables reproducible reinforcement learning for arbitrary agent harnesses, substantially improving coding-agent performance with minimal data and compute.

agent harnessesharnessed agentic RLLLM endpoint proxyretokenizationadvantage calculationloss normalizationAgent Lightning v1.0SWE-bench Verified

Abstract

Modern agents operate inside agent harnesses that manage tools, context, and control flow, making the harness a critical part of the agent system. Our original Agent Lightning introduced a disaggregated architecture that connects arbitrary agents to RL training through an LLM endpoint proxy, an approach later adopted by frameworks such as verl Uni-Agent, AReaL 2.0, slime, and Polar. We refer to this paradigm as harnessed agentic RL, where the deploy-time harness directly participates in model post-training. Harnessed agentic RL differs fundamentally from traditional agentic RL: the harness, rather than the training engine, owns the environment interaction loop, while the trainer observes only sequences of LLM request-response pairs. This introduces challenges in retokenization, sample merging, advantage calculation, loss normalization, and backend scheduling, which can substantially affect training stability and effectiveness. We present Agent Lightning v1.0, a lightweight framework for harnessed agentic RL implemented in approximately 3,500 lines of code. It supports arbitrary agent harnesses and serves as a practical testbed for studying these challenges. We evaluate it on instruction-following, search, and coding agents, and provide a complete reproducible pipeline for coding-agent RL. Using only 6K training examples and modest compute, RL improves Qwen3.5-9B on SWE-bench Verified from 41.8% to 56.4%, a 14.6-point absolute gain. We release the complete workflow and training scripts to facilitate reproducible research on harnessed agentic RL.

北京市昌平区探索星信息技术及软件开发工作室

京ICP备2026059466号
Agent Lightning v1.0: Towards Harnessed Agentic RL | TensorX