AI Paper Atlas

Research map / Reinforcement learning / Offline RL

Diffusion Models as Optimizers for Efficient Planning in Offline RL

Renming Huang, Yunqiang Pei, Guoqing Wang, Yangming Zhang, Yang Yang, Peng Wang, Heng Tao Shen

ECCV 2024 · Reinforcement learning › Offline RL · approach: Diffusion models for offline decision making · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace diffusion trajectory generation with autoregressive transformer plus few-step diffusion optimizationTrajectory Diffuser · D4RL Gym-locomotionruntime per action (s): 4.61 -> 0.44 (vs DD)Improved
replace diffusion trajectory generation with transformer plus diffusion optimizationTrajectory Diffuser · D4RL Gym-locomotionaverage normalized return: 81.8 -> 84.2 (vs DD)Improved
add diffusion-based trajectory optimization after transformer generation (ablation)Trajectory Diffuser · D4RL Gym-locomotionaverage normalized return: 79.1 -> 84.2Improved
use Transformer instead of GRU or LSTM for the feasible trajectory generator (ablation)Trajectory Diffuser · D4RL Med-Expert HalfCheetahnormalized return: 80.8 / 83.1 -> 92.2Improved
increase diffusion optimization steps from 5 to 20 or more (ablation)Trajectory Diffuser · D4RL Med-Expertaverage normalized return: 105.0 -> 98.4 (20 steps)Hurt
reduce diffusion optimization steps from 5 to 2 (ablation)Trajectory Diffuser · D4RL Med-Expertaverage normalized return: 105.0 -> 103.7No clear effect
apply Trajectory Diffuser on Adroit Pen versus Decision DiffuserTrajectory Diffuser · Adroit Penaverage normalized return: 23.9 -> 96.7Improved
apply Trajectory Diffuser on Kitchen Mixed versus Decision DiffuserTrajectory Diffuser · Franka Kitchen Mixednormalized return: 65.0 -> 54.2Hurt
use Trajectory Diffuser versus DD on Medium HalfCheetahTrajectory Diffuser · Medium HalfCheetahnormalized return: 49.1 -> 44.2Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy