AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Optimistic Exploration in Reinforcement Learning Using Symbolic Model Estimates

Sarath Sreedharan, Michael Katz

NeurIPS 2023 · Reinforcement learning › Visual and representation RL · approach: Hierarchical skills and priors · 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
bootstrap from lifted model learned on 5x5 grid (ablation)symbolic exploration with lifting · DoorKey-8x8samples: 367749.2 -> 143223.4Improved
bootstrap from lifted model learned on 5x5 grid (ablation)symbolic exploration with lifting · DoorKey-8x8time (s): 92.78 -> 52.1Improved
bootstrap from lifted model learned on 5x5 grid (ablation)symbolic exploration with lifting · DoorKey-6x6samples: 21531.8 -> 27666.0Hurt
bootstrap from lifted model learned on 5x5 grid (ablation)symbolic exploration with lifting · DoorKey-6x6time (s): 35.91 -> 42.14Hurt
bootstrap from 5x5 grid model (ablation)symbolic exploration with lifting · DoorKey-16x16completionImproved
apply bootstrapping under the original PruneModel strategy (ablation)symbolic exploration · MinigridperformanceHurt
replace original PruneModel with strategy reintroducing actions more often (ablation)symbolic exploration with bootstrapping · DoorKey-8x8samples: 260839.8 -> 143223.4Improved
add lifting generalization (ablation)symbolic exploration · planning domainsproblems solvedImproved
use R-max or SMDP exploration instead of symbolic explorationR-max, SMDP · planning and Minigridgoal reachingHurt
initialize Q values from plan found by explorationtabular Q learning · Minigridepisode rewardImproved
initialize Q values from plan on largest maptabular Q learning · largest Minigridepisode reward after 1000 episodesImproved

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