CAMP-II：用在线蒸馏解锁预训练文件操作作为智能体记忆能力
Coding Agent Memory Post-Training II: Unlocking the Memory Potential of Pre-trained File Operations for Long-Horizon Tasks via On-Policy Distillation
做智能体记忆和长任务训练的可以看看这篇。它没发明新记忆工具，而是把预训练里已有的文件操作技能直接当记忆用，再用异步在线蒸馏把教师的经验迁移过来，增益是最强 OPD 基线的 2.97 倍，还迁移到了 Terminal-Bench 2.1 和 SWE-Bench Pro V2。
一篇 arXiv 论文提出 Coding Agent Memory Post-Training II（CAMP-II），研究长时程任务中智能体如何把文件操作用作记忆。核心问题是：现有 on-policy distillation（OPD）依赖模型预训练之外的专用记忆工具和压缩协议，导致教师模型无法指导没用过的接口，学生也难以采到足够有意义的记忆动作。
为此，作者构建了 Coding Agent Memory Gym II（CAMG-II），覆盖 Shop、Coding、AutoResearch、General、Visual 五类长时程 agentic-RL 环境，每个环境提供原生任务接口之外的普通 Bash 访问和跨回合持久的工作区，让策略能在上下文被替换时通过创建、修改、检索和复用文件来充当记忆。CAMP-II 用全异步 OPD 让单个策略在五个环境上联合训练，直接从冻结教师那里学会这种基于文件的记忆行为。
在 CAMG-II 上，CAMP-II 取得了最强 OPD 基线 2.97 倍的增益，且学到的记忆行为可迁移到 Terminal-Bench 2.1 和 SWE-Bench Pro V2 两个外部基准，成功率均有提升。
Coding Agent Memory Post-Training II: Unlocking the Memory Potential of Pre-trained File Operations for Long-Horizon Tasks via On-Policy Distillation
Language-model agents increasingly tackle long-horizon tasks whose interaction histories exceed the model's active context. They must therefore decide what to write down and what to read back, and when to do either. Recent work has begun to use on-policy distillation (OPD) to make memory control part of the policy, relying on memory tools and compaction protocols that lie outside either model's pre-training. However, this weakens the distillation on both sides: the teacher cannot give useful guidance on an interface it has never used, and the student samples too few meaningful memory actions for that guidance to take effect. To study this problem, we introduce Coding Agent Memory Gym II (CAMG-II), a suite of long-horizon agentic-RL environments spanning Shop, Coding, AutoResearch, General, and Visual. Alongside each environment's native task interface, CAMG-II provides ordinary Bash access and an episode-persistent workspace, enabling the policy to create, revise, search, and reuse files as memory across context replacements. We also introduce Coding Agent Memory Post-Training II (CAMP-II), which trains a single policy jointly across all five environments with fully asynchronous OPD to learn this file-based memory behavior directly from a frozen teacher with a prior for file operations. On CAMG-II, CAMP-II obtains 2.97 times the gain of the strongest OPD baseline. The memory behavior CAMP-II learns generalizes to two external benchmarks, Terminal-Bench 2.1 and SWE-Bench Pro V2, with improved success rates.
本条情报归属 AITOP「论文」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：CAMP-II、on-policy distillation、agentic-RL、SWE-Bench Pro，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.LG」（@Lirui Luo, Kelong Mao, Xinwei Yang, Rongqing Li, Luyu Chen, Kieran Wong, Yudong Guo, Xinrui Wang, Jiayin Zhu, Gu Simiu, Sulong Xu, Cong Fang），经 AITOP 多维评分管线结构化提炼。
