L0：强化学习成为通用智能体
L0: Reinforcement Learning to Become General Agents
AI总结：
提出L0端到端训练流程与NB-Agent支架，利用可验证奖励强化学习显著提升基础模型在事实问答基准上的问题解决能力。
摘要：
训练大型语言模型（LLMs）作为多轮、长周期任务的自主智能体在可扩展性和训练效率方面仍然存在重大挑战。为了解决这个问题，我们引入了L-Zero（L0），这是一个可扩展的、端到端的通用智能体训练流程。L0具有低成本、可扩展且沙盒化的并发智能体工作池，降低了在复杂环境中应用强化学习的门槛。我们还引入了NB-Agent，即L0内部的智能体支架，它通过读取-求值-打印循环（REPL）以“代码即动作”的方式运行。我们在事实性问答基准上评估了L0。我们的实验表明，基础模型可以仅使用可验证奖励的强化学习（RLVR）来发展出稳健的问题解决技能。在Qwen2.5-7B-Instruct模型上，我们的方法将SimpleQA的准确率从30%提高到80%，将HotpotQA的准确率从22%提高到41%。我们已经开源了整个L0系统，包括我们的L0系列模型、NB-Agent、完整的训练流程以及相应的训练方案，网址为(https://github.com/cmriat/l0)。
英文摘要：
Training large language models (LLMs) to act as autonomous agents for multi-turn, long-horizon tasks remains significant challenges in scalability and training efficiency. To address this, we introduce L-Zero (L0), a scalable, end-to-end training pipeline for general-purpose agents. Featuring a low-cost, extensible, and sandboxed concurrent agent worker pool, L0 lowers the barrier for applying reinforcement learning in complex environments. We also introduce NB-Agent, the agent scaffold within L0, which operates in a "code-as-action" fashion via a Read-Eval-Print-Loop (REPL). We evaluate L0 on factuality question-answering benchmarks. Our experiments demonstrate that a base model can develop robust problem-solving skills using solely Reinforcement Learning with Verifiable Rewards (RLVR). On the Qwen2.5-7B-Instruct model, our method boosts accuracy on SimpleQA from 30 % to 80 % and on HotpotQA from 22 % to 41 %. We have open-sourced the entire L0 system, including our L0 series models, the NB-Agent, a complete training pipeline, and the corresponding training recipes on (https://github.com/cmriat/l0).
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
