#policy learning

try —

6 papers match

cs.LG2026

Multi-channel Uplift Policy Learning

Changjian Liu, Tianyu Wang, Xiaoxuan Deng +7

The paper proposes ReAlloc, a causal teacher‑student framework for allocating fixed marketing budgets across multiple e‑commerce channels, using unbiased local gradients and long‑t…

#uplift modeling#causal inference#budget allocation#multi-channel marketing
cs.AI2026

SCOPE: Supply-Chain Operations through Coupled Policies for End-to-End Coordination

Yunhao Liang, Xianqi Cao, Pujun Zhang +4

The paper introduces SCOPE, a unified policy framework that jointly optimizes assortment, source assignment, replenishment frequency, and routing decisions in supply chains, showin…

#supply chain#replenishment planning#joint optimization#logistics
cs.RO2026

DLAM: Distributional Latent Actions with Temporal Constraints

Zuojin Tang, Feifan Luo, Haoyun Liu +10

The paper introduces DLAM, a distributional latent-action model that encodes video transitions as diagonal Gaussians with temporal constraints, improving reconstruction consistency…

#latent action modeling#vision-language-action#temporal constraints#distributional dynamics
cs.LG2026

Learning Who to Treat When Treatment is Missing

Johnna Sundberg, Rayid Ghani, Eli Ben-Michael +1

The paper develops efficient estimators for policy learning when treatment assignments are missing, handling both missing-at-random and missing-completely-conditionally-at-random s…

#policy learning#missing data#causal inference#treatment effect estimation
cs.RO2026

ChunkFlow: Towards Continuity-Consistent Chunked Policy Learning

Zhao Yang, Yinan Shi, Mingyuan Yao +3

ChunkFlow introduces a seam‑aware training and execution framework for chunked robot policies that reduces boundary jitter by using deterministic overlap blending and continuity lo…

#robotic manipulation#vision-language models#policy learning#chunked actions
cs.RO2026

EDAR: Learning Environment-Dependent Action Representations for Robotic Manipulation

Yuecheng Xu, Tong Yang, Jingkai Jia +3

The paper introduces EDAR, a method that learns action representations for robotic manipulation by linking control commands with the visual effects they cause in a given environmen…

#action representation#environment-dependent#robotic manipulation#visual grounding