Mixed-Density Diffuser: Efficient Planning with Non-Uniform Temporal Resolution

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Stambaugh, Crimson, Rao, Rajesh P. N.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908960929349632
author Stambaugh, Crimson
Rao, Rajesh P. N.
author_facet Stambaugh, Crimson
Rao, Rajesh P. N.
contents Recent studies demonstrate that diffusion planners benefit from sparse-step planning over single-step planning. Training models to skip steps in their trajectories helps capture long-term dependencies without additional memory or computational cost. However, predicting excessively sparse plans degrades performance. We hypothesize this temporal density threshold is non-uniform across a planning horizon and that certain parts of a predicted trajectory should be more densely generated. We propose Mixed-Density Diffuser (MDD), a diffusion planner where the densities throughout the horizon are tunable hyperparameters. We show that MDD surpasses the SOTA Diffusion Veteran (DV) framework across the Maze2D, Franka Kitchen, and Antmaze Datasets for Deep Data-Driven Reinforcement Learning (D4RL) task domains, achieving a new SOTA on the D4RL benchmark.
format Preprint
id arxiv_https___arxiv_org_abs_2510_23026
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mixed-Density Diffuser: Efficient Planning with Non-Uniform Temporal Resolution
Stambaugh, Crimson
Rao, Rajesh P. N.
Artificial Intelligence
Robotics
Recent studies demonstrate that diffusion planners benefit from sparse-step planning over single-step planning. Training models to skip steps in their trajectories helps capture long-term dependencies without additional memory or computational cost. However, predicting excessively sparse plans degrades performance. We hypothesize this temporal density threshold is non-uniform across a planning horizon and that certain parts of a predicted trajectory should be more densely generated. We propose Mixed-Density Diffuser (MDD), a diffusion planner where the densities throughout the horizon are tunable hyperparameters. We show that MDD surpasses the SOTA Diffusion Veteran (DV) framework across the Maze2D, Franka Kitchen, and Antmaze Datasets for Deep Data-Driven Reinforcement Learning (D4RL) task domains, achieving a new SOTA on the D4RL benchmark.
title Mixed-Density Diffuser: Efficient Planning with Non-Uniform Temporal Resolution
topic Artificial Intelligence
Robotics
url https://arxiv.org/abs/2510.23026