Temporally Interpretable Differentiable Decision Trees
Temporally Interpretable Differentiable Decision Trees
Researchers introduce temporal interpretability for DDTs, improving decision-making via action chunking. Their method achieves 80% fewer parameters while matching neural policies in three domains.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S4
- code: Available Artifact link Open source S5
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
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 interpret
Open source S4
able 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.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org