AI Paper Atlas

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

Partial Identification of Policy Values under Network Interference

Ziyan Wang, Yiran Liu, Zhiheng Zhang

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Confidence intervals and OPE benchmarks

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
increase Lipschitz constant L from 0.1 to 0.5 (ablation)Partial-identification interval · Pokec (S=1)interval width: 0.022366 -> 0.101170Hurt
increase Lipschitz constant L from 0.1 to 0.5 (ablation)Partial-identification interval · Pokec (S=1)empirical coverage: 0.44 -> 0.98Improved
increase Lipschitz constant L from 0.1 to 0.5 (ablation)Partial-identification interval · Pokec (S=3)empirical coverage: 0.62 -> 1.00Improved
increase truncation depth S from 1 to 3 (ablation)Partial-identification interval · Pokec (L=0.5)interval width: 0.101170 -> 0.123289Hurt
increase truncation depth S from 1 to 3 (ablation)Partial-identification interval · Pokec (L=0.1)empirical coverage: 0.44 -> 0.38 -> 0.62Mixed
add graph-distance term to the composite metric (lambda > 0) (ablation)Partial-identification interval · Pokec, BA and SBM graphsempirical coverageImproved
use finer exposure binning K (ablation)Partial-identification interval · PokeccoverageHurt

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