AI 中文总结
该研究针对后训练导致的灾难性遗忘问题，提出让智能体在轨迹中途切换专用LoRA适配器的SLAaaT方法，实验显示其可解决新问题、自主切换策略且能力损耗低，在任务能力和token使用上优于生成子智能体的方法。
AI 中文摘要
后训练可解锁新能力并提升特定任务性能，但有时会以其他领域的灾难性遗忘为代价，这对由不同能力组成的长智能体轨迹构成问题。我们通过赋予智能体在轨迹中途切换专用LoRA适配器的工具，来拒绝这种权衡。为测试其有效性，我们构建了两个逻辑简单但需要专业化的合成编码任务。我们发现这使模型能够解决之前无法解决的问题，模型能够自主切换（并找到一种新策略，在一项任务上击败我们的人类启发式基线），与仅使用一个专用适配器的智能体相比，这带来了高达18倍的能力损耗减少。我们的方法在任务能力和token使用量上也显著优于生成子智能体的方法。
英文摘要
Post-training can unlock new capabilities and improve performance on specialized tasks, but sometimes at the cost of catastrophic forgetting in other domains. This poses a problem in long agent trajectories that compose different capabilities. We reject this tradeoff by giving an agent a tool to switch between specialized LoRA adapters mid-trace. To test its effectiveness, we compose two synthetic coding tasks that are logically simple but require specialization. We find that this allows the model to solve problems it previously could not, that the model is able to switch autonomously (and find a new strategy that beats our human heuristic baseline on one task), and that this incurs an up to an 18x reduction in capability tax compared to an agent using only one specialized adapter. Our approach also substantially outperforms spawning subagents in both task capabilities (solving 4 of our hardest tasks versus none) and token usage (46.1x fewer tokens in some scenarios).
