AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Q-Learning with Adjoint Matching

Qiyang Li, Sergey Levine

ICLR 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Model-based RL with guarantees · 22 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
use the lean adjoint state instead of the basic adjoint matching objective (backprop through SDE) (ablation)QAM · OGBench offline RL (50 tasks)success rate (%): 35 -> 44Improved
use the critic action gradient instead of only its value (advantage weighting)QAM · OGBench offline RL (50 tasks)success rate (%): 8 -> 44Improved
combine QAM with FQL one-step distillation (QAM-FQL)QAM · OGBench offline RL (50 tasks)success rate (%): 44 -> 45Improved
combine QAM with a FEdit edit policy (QAM-EDIT)QAM · OGBench offline RL (50 tasks)success rate (%): 44 -> 46Improved
add SOC formulation with optimal-policy guarantee vs FBRAC direct backpropBAM · OGBench offline RL (50 tasks)success rate (%): 11 -> 35Improved
compare BAM (backprop through SDE) with FQL distilled one-step policyBAM · OGBench offline RL (50 tasks)success rate (%): 35 vs 36No clear effect
enable gradient clipping (ablation)QAM-EDIT · OGBench offline RL (50 tasks)success rate (%)Improved
use large critic ensemble K=10 (ablation)QAM-EDIT · OGBench offline RL (50 tasks)success rate (%)Improved
increase flow steps T beyond 10 (ablation)QAM-EDIT · OGBench offline RL (50 tasks)success rate (%)No clear effect
tune temperature tau (ablation)QAM-EDIT · OGBench offline RL (50 tasks)success rate (%)Improved
train on noisy datasets (cube-triple-noisy) instead of playQAM-EDIT · OGBench cube-triple-noisysuccess rate (%)Mixed
train on stitch datasets instead of navigate for locomotionQAM-EDIT · OGBench stitch navigationsuccess rate (%)No clear effect

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