arXiv：cs.LG· Eisuke Hirota, Aarav Sane, Rohan Paleja·· 4 小时前AI 评分34
可微决策树迎来时间维度可解释性：动作分块让 DDT 策略更易理解
Temporally Interpretable Differentiable Decision Trees
AI 导读
研究提出"时间可解释性"概念，通过动作分块（action chunking）让可微决策树（DDT）适配序贯决策任务，并给出两种策略梯度算法及一种信息论树重构算法。在四个仿真环境中，由蒸馏后的动作分块策略热启动的 DDT，在其中三个环境里达到神经网络策略水平，参数量最多减少 80%。
正文
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 this https URL.
| Subjects: | Machine Learning (cs.LG); Robotics (cs.RO) |
| Cite as: | arXiv:2610.10367 [cs.LG] |
| (or arXiv:2610.10367v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.10367 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Eisuke Hirota [view email]
[v1]
Wed, 7 Oct 2026 16:35:17 UTC (293 KB)
来源：arXiv：cs.LG · arxiv.org