AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learning

Luisa Zintgraf, Kyriacos Shiarlis, Maximilian Igl, Sebastian Schulze, Yarin Gal, Katja Hofmann, Shimon Whiteson

ICLR 2020 · Reinforcement learning › Visual and representation RL · approach: Task structure and curricula for meta-RL · 94 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 a stochastic task embedding and infer its posterior from experiencevariBAD · Gridworldaverage returnImproved
Use a stochastic task embedding and infer its posterior from experiencevariBAD · MuJoCo continuous control tasksonline returnImproved
Condition action selection on the task posteriorvariBAD · Gridworldexploration efficiencyImproved
Decode future as well as past transitions and rewardsvariBAD · Meta-RL tasksinference about unseen statesImproved
Remove the decoder and VAE objective, leaving a recurrent policy (ablation)variBAD · Meta-RLtask representationHurt
Train with a reward decoder only rather than also a dynamics decoder (ablation)variBAD · MuJoCo continuous control tasksperformanceImproved
Do not backpropagate the RL loss through the encoder (ablation)variBAD · Meta-RL trainingtraining speedImproved
Do not backpropagate the RL loss through the encoder (ablation)variBAD · Meta-RL trainingpolicy learningNo clear effect
Replace the learned stochastic latent belief and decoder with an RL2-style recurrent policy (ablation)variBAD · Gridworldmatching Bayes-optimal solutions across seedsMixed

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