AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Discovering Hierarchical Achievements in Reinforcement Learning via Contrastive Learning

Seungyong Moon, Junyoung Yeom, Bumsoo Park, Hyun Oh Song

NeurIPS 2023 · 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 achievement distillation (contrastive auxiliary learning)PPO · Crafter 1M stepsscore (%): 15.60 -> 21.79Improved
Add intra-trajectory achievement prediction (ablation)PPO · Crafter 1M stepsscore (%): 15.60 -> 19.02Improved
Add cross-trajectory achievement matching (ablation)PPO + intra-trajectory prediction · Crafter 1M stepsscore (%): 19.02 -> 20.36Improved
Add memory (ablation)PPO + prediction + matching · Crafter 1M stepsscore (%): 20.36 -> 21.79Improved
Shrink model to 1M parameters (channels [16, 32, 32], hidden 256)Ours · Crafter 1M stepsscore (%): 21.79 -> 17.07Hurt
Add achievement distillationPPO · Crafter 1M stepsclassification accuracy (%): 73.6Improved
Add achievement distillationQR-DQN · Crafter 1M stepsscore (%): 4.14 -> 8.07Improved
Add achievement distillationPPO · Procgen Heistscore: 29.6 -> 71.0Improved
Add achievement distillationPPO · MiniGridscore: 3.33 -> 8.04Improved
Use PPO with recent implementation practices instead of DreamerV3PPO · Crafter 1M stepsscore (%): 14.77 -> 15.60Improved
Use PPO instead of LSTM-SPCNN (recurrent object-centric network)PPO · Crafter 1M stepsscore (%): 11.67 -> 15.60Improved
Our method (9M params) versus MuZero + SPR with pre-trainingachievement distillation · Crafter 1M stepsscore (%): 16.4 -> 21.79Improved

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