AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

There Is No Turning Back: A Self-Supervised Approach for Reversibility-Aware Reinforcement Learning

Nathan Grinsztajn, Johan Ferret, Olivier Pietquin, Philippe Preux, Matthieu Geist

NeurIPS 2021 · Reinforcement learning › Visual and representation RL · approach: In-context and adaptive RL · 1 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
add reversibility-aware exploration (RAE) intrinsic reward penalizing irreversibility, with no extrinsic rewardPPO · reward-free Cartpoleepisode scoreImproved
add RAE to the agentPPO · Turfpath chosen / irreversible side-effectsImproved
add RAE to IMPALAIMPALA · Sokoban 1k levelslevels solvedImproved
augment a random policy with reversibility-aware control (RAC)random policy · Cartpole+score: 1152 -> 50kImproved
add RAC action filtering with threshold 0.2PPO · Turfirreversible side-effects during trainingImproved
add RAC action filteringPPO · Turflearning speedHurt
vary RAC reversibility threshold: too low versus too high (ablation)RAC · Cartpole+, Turfsafety vs task performanceMixed

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