AI Paper Atlas

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

XQC: Well-conditioned Optimization Accelerates Deep Reinforcement Learning

Daniel Palenicek, Florian Vogt, Joe Watson, Ingmar Posner, Jan Peters

ICLR 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Regularized scalable deep value learning · 8 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 batch normalization instead of layer normalization in the critic (ablation)XQC critic · DMC dog-trot Hessian condition numbercondition number: orders of magnitude smallerImproved
use categorical cross-entropy critic loss instead of MSE (ablation)XQC critic · DMC dog-trot Hessian condition numbercondition numberImproved
combine BN, weight normalization and CE loss (ablation)SAC critic · DMC dog-trotreturn at 1M steps / conditioningImproved
replace BN with LN in XQC (ablation)XQC · 55 proprioceptive tasksaggregate normalized returnHurt
halve learning rate for LN variant (ablation)XQC with LN · 55 proprioceptive tasksaggregate normalized returnHurt
replace CE critic loss with MSE (ablation)XQC · 55 proprioceptive tasksaggregate normalized returnHurt
remove weight normalization (ablation)XQC · 55 proprioceptive tasksaggregate normalized returnHurt
add skip connections at investigated depths (ablation)XQC · depth scalingperformanceNo clear effect
vary batch size (128, 512 vs 256), add weight decay, action repetition (ablation)XQC · ablationsperformanceNo clear effect
increase update-to-data ratio (ablation)XQC · UTD scalingAUC of IQM normalized returnImproved
remove weight normalization (growing parameter norms) (ablation)XQC · plasticity analysiseffective learning rateHurt

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