AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Offline Multi-Objective Bandits: From Logged Data to Pareto-Optimal Policies

Ji Cheng, Song Lai, Shunyu Yao, Bo Xue

AAAI 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Tree search and policy optimization

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 pessimistic lower-confidence-bound action selection instead of greedy decisions (ablation)OffMOB (neural reward models) · synthetic nonlinear multi-objective contextual bandits (m=3, d=20, K=30)Tchebycheff sub-optimalityImproved
add pessimism to a fixed-feature neural linear model (MONeuralLinLCB vs MONeuralLinGreedy) (ablation)MONeuralLinLCB · synthetic nonlinear multi-objective contextual banditsTchebycheff sub-optimalityNo clear effect
use trainable neural networks in NTK space instead of linear, kernel or fixed-feature modelsOffMOB · synthetic nonlinear multi-objective contextual banditsTchebycheff sub-optimalityImproved
increase observation noise sigma from 0.01 to 0.5 (ablation)OffMOB and baselines · quadratic reward, sigma 0.01, 0.1, 0.5Tchebycheff sub-optimalityHurt
use OffMOB under higher noise relative to next-best methodsOffMOB · quadratic reward, sigma = 0.5Tchebycheff sub-optimalityImproved

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