AI Paper Atlas

Research map / Reinforcement learning / Offline RL

Exploit Reward Shifting in Value-Based Deep-RL: Optimistic Curiosity-Based Exploration and Conservative Exploitation via Linear Reward Shaping

Hao Sun, Lei Han, Rui Yang, Xiaoteng Ma, Jian Guo, Bolei Zhou

NeurIPS 2022 · Reinforcement learning › Offline RL · approach: Offline RL sample complexity theory · 10 citations (OpenAlex)

Read the paper

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 a positive reward shiftBCQ and CQL · offline RL (D4RL-style tasks incl. Hopper)returnImproved
add a negative reward shift (ablation)BCQ and CQL · offline RLreturnHurt
add Random Reward Shift (3 Q-networks with shifts +0.5, 0, -0.5)TD3 · MuJoCo continuous control (5 envs)returnImproved
set all reward shifts to 0 (Ensemble TD3 vs RRS) (ablation)RRS · MuJoCo continuous controlreturnImproved
add a -0.5 reward shiftDQN · MountainCar and MiniGridexploration performanceImproved
use positive intrinsic reward (vanilla RND) with DQNDQN · MountainCar and MiniGridexploration performanceHurt
make RND intrinsic reward negative by subtracting a constantRND with DQN · MountainCar and MiniGridexploration performanceImproved
increase negative shift magnitude from -1.0 to -1.5 (ablation)RND with DQN · MountainCar and MiniGridexploration performance: -1.0 -> -1.5Improved

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