信念外学习
Off-Belief Learning
摘要：
Dec-POMDPs中的标准问题设置是自我博弈，其目标是找到一组最佳的策略。通过自我游戏学习的策略可能会采用任意的约定，并隐含地依赖于基于对其他代理行为的脆弱假设的多步推理，因此在测试时与人类或独立训练的代理配对时会失败。为了解决这个问题，我们提出了信念偏离学习（OBL）。在每个时间步，OBL代理遵循一个策略$π_1$，该策略在假设过去的动作由给定的固定策略（$π_0$）执行，但假设未来的动作将由$π_1$执行的情况下进行优化。当$π_0$是均匀随机的，OBL收敛到一个最优的政策，不依赖于推理的基础上其他代理的行为（最优接地政策）。OBL可以在层次结构中迭代，其中一个级别的最佳策略成为下一个级别的输入，从而以受控的方式引入多级认知推理。与现有的方法，这可能会收敛到任何均衡的政策，OBL收敛到一个独特的政策，使其适合于zero-shot协调（ZSC）。OBL可以通过虚构的转换机制扩展到高维设置，并且在玩具设置和基准人类AI & ZSC问题Hanabi中表现出强大的性能。
英文摘要：
The standard problem setting in Dec-POMDPs is self-play, where the goal is to find a set of policies that play optimally together. Policies learned through self-play may adopt arbitrary conventions and implicitly rely on multi-step reasoning based on fragile assumptions about other agents' actions and thus fail when paired with humans or independently trained agents at test time. To address this, we present off-belief learning (OBL). At each timestep OBL agents follow a policy $π_1$ that is optimized assuming past actions were taken by a given, fixed policy ($π_0$), but assuming that future actions will be taken by $π_1$. When $π_0$ is uniform random, OBL converges to an optimal policy that does not rely on inferences based on other agents' behavior (an optimal grounded policy). OBL can be iterated in a hierarchy, where the optimal policy from one level becomes the input to the next, thereby introducing multi-level cognitive reasoning in a controlled manner. Unlike existing approaches, which may converge to any equilibrium policy, OBL converges to a unique policy, making it suitable for zero-shot coordination (ZSC). OBL can be scaled to high-dimensional settings with a fictitious transition mechanism and shows strong performance in both a toy-setting and the benchmark human-AI & ZSC problem Hanabi.
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。
