AI Paper Atlas

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

Intentional Updates for Streaming Reinforcement Learning

Arsalan Sharifnassab, Mohamed Elsayed, Kris De Asis, A. Rupam Mahmood, Richard S Sutton

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Regularized scalable deep value learning · 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
Replace ObGD update with intentional updates (StreamAC -> Intentional AC)streaming actor-critic · MuJoCo Ant-v4average return: 4898 -> 5513Improved
Replace ObGD update with intentional updatesstreaming actor-critic · MuJoCo Humanoid-v4average return: 4920 -> 6227Improved
Replace ObGD update with intentional updatesstreaming actor-critic · MuJoCo Walker2d-v4average return: 3590 -> 3515No clear effect
Intentional TD instead of StreamTD(lambda)critic under fixed policy · MuJoCo value predictionnormalized RMS error: StreamTD up to 2.54x error of Intentional TDImproved
Add trace-aware sigma term to step size (vs naive eta/<rho z, z>) (ablation)Intentional AC · MuJoCoaverage returnImproved
Add diagonal RMSProp scaling (ablation)Intentional AC · MuJoCoaverage returnImproved
Add adaptive delta clipping (ablation)Intentional AC · MuJoCoaverage returnNo clear effect
Keep SparseInit (ablation)Intentional AC · MuJoCo Humanoid-v4average return: 2503 -> 6227Improved
Keep SparseInit (ablation)StreamAC · MuJoCo Humanoid-v4average return: 610 -> 4920Improved
Keep LayerNorm (ablation)Intentional AC · MuJoCo Ant-v4average return: 2323 -> 5513Improved
Intentional scaling vs alpha=1 on effective update ratio (ablation)Intentional AC critic · MuJoCoeffective update ratio: 2.99 -> 1.84Improved
Intentional scaling vs alpha=1 on effective update ratio (ablation)Intentional AC policy · MuJoCoeffective update ratio: 1.83 -> 2.61Hurt

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