# policy learning

4 related articles

Subscribe to updates

arXiv Oct 29, 2026
Training logs know what deployment cannot: runtime-confounded RL gets doubly robust evaluation and a testable price
When deployment lacks key variables that training logs record in full, runtime confounding can push a reinforcement-learning policy off course over time. The work places the runtime restriction only in the numerator of an importance ratio and estimates nuisance functions from the full data, yielding an asymptotically normal off-policy estimator that attains the semiparametric efficiency bound, a policy-learning regret bound that differs from the full-information optimum by an irreducible runtime gap, and a sample-splitting test for that gap.
When deployment lacks key variables that training logs record in full, runtime confounding can push a reinforcement-learning policy off course over time. The work places the runtime restriction only in the numerator of an importance ratio and estimates nuisance functions from the full data, yielding an asymptotically normal off-policy estimator that attains the semiparametric efficiency bound, a policy-learning regret bound that differs from the full-information optimum by an irreducible runtime gap, and a sample-splitting test for that gap.
Johns Hopkins University New York University Cornell University
Raphael Kim（First author）
Reinforcement Learning policy learning Off-policy evaluation Runtime confounding
arXiv Oct 27, 2026
Not every document question needs an agent: RouteDRAG sends 79.3% of queries to agentic execution and gains 3.47 points on average
In multimodal long-document question answering, agentic retrieval is not always worth its extra computation and can even hurt answer accuracy. RouteDRAG makes both whether to invoke an agent and how it then executes into learned choices: a graph-backed index organizes cross-page, table and figure content, path routing predicts the benefit of an agent over shorter paths, and dynamic execution controls tool use and termination. Across two benchmarks it assigns only 79.3% of queries to an agent while averaging 3.47 percentage points higher accuracy than shorter-path-only execution, and it also surpasses uniform agent execution with fewer answering calls.
In multimodal long-document question answering, agentic retrieval is not always worth its extra computation and can even hurt answer accuracy. RouteDRAG makes both whether to invoke an agent and how it then executes into learned choices: a graph-backed index organizes cross-page, table and figure content, path routing predicts the benefit of an agent over shorter paths, and dynamic execution controls tool use and termination. Across two benchmarks it assigns only 79.3% of queries to an agent while averaging 3.47 percentage points higher accuracy than shorter-path-only execution, and it also surpasses uniform agent execution with fewer answering calls.
Zhejiang University Fudan University Chinese Academy of Sciences National University of Defense Technology Westlake University
Zekai Chen（First author）
RouteDRAG Agentic Retrieval-Augmented Generation policy learning
arXiv Oct 26, 2026
Regional Policy and Solo Founders Can Now Hear Each Other: A Mediating Agent Lifts the Government Objective 42.4% Above the Eight-Baseline Mean
Chinese regions use separate support policies to compete for one-person companies, solo founders who work through digital platforms and AI agents and can compare locations, and the two sides used to act on their own separate estimates. PolicyMediator turns that into a cognition-constrained leader-multi-follower Stackelberg framework: a mediator agent explains verified policies to the companies to build awareness and trust, then reports their needs and unresolved concerns to governments for policy learning. Across six-region simulations in China with five OPC types, five scenarios and three simulator backbones, it ranks first in all 15 combinations, with the aggregate government objective 42.4% above the eight-baseline mean.
Chinese regions use separate support policies to compete for one-person companies, solo founders who work through digital platforms and AI agents and can compare locations, and the two sides used to act on their own separate estimates. PolicyMediator turns that into a cognition-constrained leader-multi-follower Stackelberg framework: a mediator agent explains verified policies to the companies to build awareness and trust, then reports their needs and unresolved concerns to governments for policy learning. Across six-region simulations in China with five OPC types, five scenarios and three simulator backbones, it ranks first in all 15 combinations, with the aggregate government objective 42.4% above the eight-baseline mean.
Northwestern Polytechnical University
Jie Li（First author）
PolicyMediator leader-multi-follower Stackelberg framework with bilateral mediation policy learning
arXiv Oct 14, 2026
Stop copying fingertip drafts: a contact-aware controller lifts real-world teleoperation success from 29.2% to 78.3%
DexFLEX treats upstream fingertip-motion drafts as evidence about intent rather than trajectories to copy, raising real-world teleoperation success from 29.2% to 78.3% with no retraining.
Weikang Wan（First author）
DexFLEX contact-aware foundation controller policy learning

You’re all caught up

Training logs know what deployment cannot: runtime-confounded RL gets doubly robust evaluation and a testable price

Not every document question needs an agent: RouteDRAG sends 79.3% of queries to agentic execution and gains 3.47 points on average

Regional Policy and Solo Founders Can Now Hear Each Other: A Mediating Agent Lifts the Government Objective 42.4% Above the Eight-Baseline Mean

Stop copying fingertip drafts: a contact-aware controller lifts real-world teleoperation success from 29.2% to 78.3%