AI Paper Atlas

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

Q-Flow: Stable and Expressive Reinforcement Learning with Flow-based Policy

JaeHyeok Doo, Byeongguk Jeon, Seonghyeon Ye, Kimin Lee, Minjoon Seo

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Model-based RL with guarantees · 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
use Q-Flow intermediate-value gradient matching instead of FQL one-step distillationFQL · OGBench standard setting averageaverage score: 43.8 -> 54.4Improved
use Q-Flow instead of FQL on antmaze-giantFQL · antmaze-giant (standard)success rate (%): 9 -> 40Improved
use Q-Flow instead of FQL on puzzle-3x3FQL · puzzle-3x3 (standard)success rate (%): 30 -> 49Improved
use Q-Flow instead of QAM-E in advanced settingQAM-E · OGBench advanced setting averageaverage score: 44.8 -> 47.5Improved
use Q-Flow instead of QAM-E on puzzle-4x4-sparseQAM-E · puzzle-4x4-sparse (advanced)success rate (%): 36 -> 0Hurt
use Q-Flow instead of QAM-E on cube-doubleQAM-E · cube-double (advanced)success rate (%): 65 -> 38Hurt
add intermediate value gradient matching in place of BPTT through the ODE solver (ablation)Q-Flow with learned intermediate value · 5 OGBench default tasks (standard)success rateImproved
use flow-consistent intermediate value instead of outer Q function guidance (ablation)Q-Flow gradient matching · 5 OGBench default tasks (standard)success rateImproved
replace FBRAC BPTT with Q-Flow in trainingFBRAC · Puzzle-4x4 training cost, 50 flow stepstraining time per step: approximately 2x fasterImproved
use Q-Flow instead of FQL during online fine-tuningFQL · 5 OGBench default tasks (offline-to-online)average success: +23% pointsImproved

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