AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Dynamics Generalisation in Reinforcement Learning via Adaptive Context-Aware Policies

Michael Beukman, Devon Jarvis, Richard Klein, Steven James, Benjamin Rosman

NeurIPS 2023 · Reinforcement learning › Visual and representation RL · approach: Curriculum and environment design · 1 citations (OpenAlex)

Read the paper

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
ignore context (Unaware) instead of giving the policy the contextSAC policy · ODE, train {-5,-1,1,5}, eval [-10,10]average evaluation rewardHurt
use Decision Adapter instead of Concat, cGate, FLAPSAC policy · ODE 1D contextaverage evaluation rewardImproved
use Decision Adapter instead of baselines when extrapolatingSAC policy · ODE extrapolation contextsaverage evaluation rewardImproved
use Decision Adapter instead of Concat on 2D context ODESAC policy · ODE 2D contextaverage evaluation rewardImproved
add context to the policy (Concat/cGate/Adapter) versus UnawareSAC policy · CartPolesteps to reach reward above 400: 600 -> 100Improved
add distractor context variables (0 to 100)Concat and cGate · CartPoleaverage evaluation rewardHurt
add distractor context variables (0 to 100)Decision Adapter · CartPoleaverage evaluation rewardNo clear effect
sample non-fixed Gaussian distractors (train N(1,0.2), eval N(0,0.2))Decision Adapter vs Concat and cGate · CartPoleaverage evaluation rewardImproved
add distractor dimensions with different train and test valuesDecision Adapter vs Concat and cGate · Mujoco Antaverage evaluation rewardImproved
use hidden width 8 or no hidden layers in the adapter instead of bottleneck 32 (ablation)Decision Adapter · ODEaverage evaluation rewardHurt
use larger adapter hidden layer of 256 or multiple layers (ablation)Decision Adapter · ODEaverage evaluation rewardNo clear effect
use hypernetwork chunking instead of non-chunking hypernetworks (ablation)Decision Adapter · ODEaverage evaluation rewardImproved

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