AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Dynamical Distance Learning for Semi-Supervised and Unsupervised Skill Discovery

Kristian Hartikainen, Xinyang Geng, Tuomas Haarnoja, Sergey Levine

ICLR 2020 · Reinforcement learning › Visual and representation RL · approach: Reward shaping for sparse rewards · 3 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 supervised regression on on-policy trajectory pairs to learn dynamical distancesDDL · DClaw valve turning (state and vision)task successImproved
Use cumulative distance as the policy objective instead of greedily minimizing distance (ablation)DDL · DClaw valve turning (state and vision)task learningImproved
Replace supervised distance regression with temporal-difference Q-learning (ablation)DDL · DClaw valve turning (state and vision)task learningHurt
Use hindsight experience replay with preference goals instead of dynamical-distance rewards (ablation)DDLfP · DClaw valve turning (state and vision)task learningMixed
Use dynamical-distance rewards rather than a per-step reward of -1 until the goalGoal-reaching policy learning · DClaw valve turning (state and vision)task learningImproved
Use 100 preference queries with DDLfP rather than the prior preference-learning method's 750 queriesDDLfP · MuJoCo locomotion taskstrue task returnMixed
Select goals at the greatest estimated dynamical distance from the initial stateDDLUS · Hopper-v3, HalfCheetah-v3, Ant-v3distance traveledImproved
Select goals at the greatest estimated dynamical distance from the initial stateDDLUS · InvertedDoublePendulum-v2task performanceImproved
Use preference queries to specify goals instead of manually designed rewards or goal examplesDDLfP · Real-world DClaw valve turningvalve rotationImproved

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