外观
论文arXiv 2610.1036721 天前
具有时间可解释性的可微分决策树
原文链接：Temporally Interpretable Differentiable Decision Trees（arXiv 2610.10367 · 发布日期：2026-10-07）
Temporally Interpretable Differentiable Decision Trees
针对DDT单步树行为与人类多步规划不匹配，提出时间可解释性与动作分块方法，并在四个仿真环境验证。
核心要点
- 现有DDT虽保留可微策略和树形可视化，但不适合序贯决策中的多步规划理解。
- 论文提出时间可解释性，引入动作分块，并设计两种结合动作分块的策略梯度算法。
- 为保持树的参数效率，作者提出信息论树重构算法，在训练过程中修改树结构。
- 在四个仿真环境中，蒸馏策略热启动的动作分块DDT效果最好；代码已开源。
验证范围与局限：验证范围为四个仿真环境，摘要未涉及真实机器人实验。
英文摘要 (Abstract)
Interpretability offers a solution to safe autonomy by providing transparency into an agent's underlying decision-making model. Within sequential-decision making tasks, differentiable decision trees (DDTs) are one approach to such interpretability, maintaining automatic-differentiable policies while providing humans with a discrete tree-based visualization. Nonetheless, current implementations of DDTs are not well-suited for sequential-decision making domains, as there exists an inherent mismatch between a tree's single-timestep behavior and a human's multi-timestep planning. Our work thus introduces time as a new dimension of interpretability, coined as temporal interpretability, and demonstrates how temporal abstractions via action chunking improve it. We achieve this by first introducing two novel policy gradient algorithms that incorporate action chunking. Additionally, to maintain parameter-efficient trees, we develop an information-theoretic tree restructuring algorithm that modifies the tree during training. Across four simulation environments, we find that warm-starting action chunked DDTs from a distilled action chunked policy is the most effective way to obtain temporally interpretable trees: they match neural network policies in three of the four domains while using up to 80$%$ fewer parameters. Our code is available at https://github.com/ei5uke/temp-interp.
作者：Eisuke Hirota, Aarav G Sane, Rohan Paleja
来源
本页事实对应的公开页面，以原文为准。