AI Paper Atlas

Research map / Reinforcement learning / Offline RL

BiTrajDiff: Bidirectional Trajectory Generation with Diffusion Models for Offline Reinforcement Learning

Yunpeng Qing, Yixiao Chi, Shuo Chen, Shunyu Liu, Kexuan Zhou, Sixu Lin, Litao Liu, Changqing Zou

ICML 2026 · Reinforcement learning › Offline RL · approach: Simple practical offline RL · 0 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
add backward-history generation to forward-only generation (bidirectional) (ablation)BiTrajDiff + IQL · D4RL locomotion averagenormalized return: 80.0 -> 85.5Improved
add forward-future generation to backward-only generation (bidirectional) (ablation)BiTrajDiff + IQL · D4RL locomotion averagenormalized return: 82.1 -> 85.5Improved
add backward-history generation to forward-only generation (bidirectional) (ablation)BiTrajDiff + TD3BC · D4RL locomotion averagenormalized return: 78.3 -> 83.5Improved
add forward-future generation to backward-only generation (bidirectional) (ablation)BiTrajDiff + TD3BC · D4RL locomotion averagenormalized return: 82.7 -> 83.5Improved
add the greedy reward filter on top of the OOD filter (ablation)BiTrajDiff + IQL/TD3BC · D4RL learning curvesreturnImproved
add the OOD (isolation forest) filter (ablation)BiTrajDiff + IQL/TD3BC · D4RL learning curvestraining stabilityImproved
raise augmented data ratio sigma to 30%-50% (ablation)BiTrajDiff · walker2d-medium-replayreturnImproved
raise augmented data ratio sigma to 100% (ablation)BiTrajDiff · walker2d-medium-replayreturn stabilityHurt
increase n of the n-step TD estimator (ablation)IQL/TD3BC with DA · D4RL n-step TDreturnHurt
replace baseline IQL with BiTrajDiff augmentationIQL · D4RL Maze averagenormalized return: 51.2 -> 69.7Improved
replace baseline TD3BC with BiTrajDiff augmentationTD3BC · D4RL Kitchen averagenormalized return: 6.9 -> 15.1Improved
replace baseline DT with BiTrajDiff augmentationDT · D4RL locomotion averagenormalized return: 67.7 -> 71.8Improved

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