AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Learning Options via Compression

Yiding Jiang, Evan Zheran Liu, Benjamin Eysenbach, J Zico Kolter, Chelsea Finn

NeurIPS 2022 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Gradient estimation for discrete variables · 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
Use trajectory compression to learn options instead of maximizing likelihood aloneOffline skill learning · Grid world demonstrationsF1: 0.32 -> 0.92Improved
Add minimum skill length constraint (ablation)DDO · Grid world demonstrationsF1: 0.32 -> 0.35Improved
Add minimum skill length constraint (ablation)VTA · Grid world demonstrationsF1: 0.32 -> 0.36Improved
Add minimum skill length constraint (ablation)VTA and DDO · Grid world demonstrationsBoundary qualityMixed
Use LOVE skills for downstream explorationRL agents learning new grid-world tasks · Grid world, sparse reward and N_pick = 5ReturnImproved
Use DDO skills with minimum skill length (ablation)RL agents learning new grid-world tasks · Grid world, dense reward and N_pick = 5Timesteps to solve: over 8× more timestepsHurt
Use behavior-cloned demonstrationsLow-level policy · Grid world, N_pick = 3 with dense rewardsReturnImproved
Use behavior-cloned demonstrationsLow-level policy · Other grid-world settingsReturnNo clear effect
Use VTA skills for downstream learningRL agents learning new tasks · Grid worldReturnNo clear effect
Set compression weight λ to different fixed values (ablation)LOVE · Simple Colors and Conditional ColorsBoundary F1: Simple: $.67\pm.27$ to $.95\pm.06$; Cond.: $.82 \pm .08$ to $.92 \pm .02$Mixed
Tune compression weight λ with dual gradient descent (ablation)LOVE · Simple Colors and Conditional ColorsBoundary F1: Simple: $.99\pm.00$; Cond.: $.99 \pm .00$Improved
Enforce minimum skill duration during training (ablation)LOVE · Skill optimizationOptimization stabilityImproved

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