AI Paper Atlas

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

Iterative Amortized Policy Optimization

Joseph Marino, Alexandre Piché, Alessandro Davide Ialongo, Yisong Yue

NeurIPS 2021 · Reinforcement learning › Policy gradient and actor-critic · approach: Regularized scalable deep value learning · 2 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 direct amortization with iterative amortization (5 iterations)SAC policy · MuJoCo continuous controlcumulative rewardImproved
increase policy optimization iterations per step during training from 1 to 2 to 5 (ablation)Iterative amortization · HalfCheetah-v2performance and amortization gapImproved
add normalizing-flow multi-modal policy to direct amortization (ablation)Direct amortization · Hopper-v2, Walker2d-v2cumulative rewardImproved
add normalizing-flow multi-modal policy to direct amortization (ablation)Direct amortization · HalfCheetah-v2cumulative rewardNo clear effect
run 5 additional amortized iterations at evaluation (ablation)Iterative amortization trained with 5 iterations · MuJoCo amortization gapestimated amortization gapImproved
deploy more fully optimized policy at evaluation (ablation)Iterative amortization · MuJoCo environmentsevaluation returnNo clear effect
compare iterative amortization with Adam and CEM optimizersPolicy optimization · Variational objective, 100 states per seediterations to reach objectiveImproved
transfer iterative optimizer from model-free to model-based value estimates without retrainingIterative amortization · HalfCheetah-v2cumulative rewardImproved
iterative vs direct amortization on simple pendulum tasksSAC policy · InvertedPendulum-v2, InvertedDoublePendulum-v2cumulative rewardNo 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