AI Paper Atlas

Research map / Reinforcement learning / Offline RL

Moment Matching Q-Learning

Yiyan Liang Edgar, Sifei Liu, Weitong Zhang

ICML 2026 · Reinforcement learning › Offline RL · approach: Diffusion models for offline decision making · 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 online fine-tuning after offline training (halfcheetah-medium-replay)MoMa QL · D4RL Gym offline-to-onlinenormalized score: 63.3 -> 80.9Improved
add online fine-tuning (walker2d-medium-replay)MoMa QL · D4RL Gym offline-to-onlinenormalized score: 104.3 -> 117.9Improved
add online fine-tuning (hopper-medium-replay)MoMa QL · D4RL Gym offline-to-onlinenormalized score: 106.5 -> 113.5Improved
add online fine-tuning (halfcheetah-medium)MoMa QL · D4RL Gym offline-to-onlinenormalized score: 72.6 -> 83.1Improved
add online fine-tuning on expert datasetsMoMa QL · D4RL Gym offline-to-online (expert)normalized score: 106.9 -> 107.1No clear effect
vary Q-learning weight eta in [0.5, 1.0] (ablation)MoMa QL · D4RLnormalized scoreNo clear effect
increase sampling steps N beyond 2 (ablation)MoMa QL · D4RLnormalized scoreNo clear effect
vary MMD kernel parameters a, b in [1, 8] (ablation)MoMa QL · D4RLnormalized scoreNo clear effect
vary noise schedule parameters p_mean and p_std (ablation)MoMa QL · D4RLnormalized scoreNo clear effect
increase sampling steps from 1 to 16 (ablation)MoMa QL · Gym training timeseconds per 1k steps: 17.70 -> 72.56Hurt
use moment-matching MMD objective instead of diffusion (Diffusion-BC)Diffusion-BC · D4RL Gym (behavior cloning) averagenormalized score: 76.3 -> 89.8Improved
use MoMa QL instead of Diffusion-QL on HopperDiffusion-QL · D4RL Gym (offline RL) hoppernormalized score: 101.0 -> 99.9Hurt

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