AI Paper Atlas

Research map / Reinforcement learning / Offline RL

Learning to Influence Human Behavior with Offline Reinforcement Learning

Joey Hong, Sergey Levine, Anca Dragan

NeurIPS 2023 · Reinforcement learning › Offline RL · approach: Offline RL sample complexity theory · 6 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 offline RL (CQL) instead of BCBC · Overcooked Asymmetric Advantages and Forced Coordination, standard and new tasksmean reward with human partnersImproved
use offline RL instead of filtered BCFiltered BC · Overcooked new tasksmean reward with human partnersImproved
use offline RL instead of Fictitious Co-PlayFCP · Overcooked new tasksmean reward with human partnersImproved
condition on human behavior via memory of history (memory offline RL) instead of plain offline RL (ablation)Offline RL · Open Asymmetric Advantages and Counter Circuitmean reward with human partnersImproved
learn explicit latent strategy representation (latent offline RL) instead of memory offline RL (ablation)Memory offline RL · Open Asymmetric Advantages and Counter Circuitmean reward with human partnersImproved
use latent offline RL instead of offline LILIOffline LILI · Open Asymmetric Advantages and Counter Circuitmean reward with human partnersImproved
pair agents with scripted unchangeable partner instead of human (ablation)All agents · Open Asymmetric Advantages and Counter Circuitmean rewardNo clear effect

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