AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

On the Tension Between Optimality and Adversarial Robustness in Policy Optimization

Haoran Li, Jiayu Lv, Congying Han, Zicheng Zhang, Anqi Li, Yan Liu, Tiande Guo, Nan Jiang

ICLR 2026 · Adversarial robustness and security › Certified and empirical robustness · approach: Provable robust learning guarantees · 0 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
train with ARPO (strongest adversary) instead of SPOPPO · MuJoCo (Hopper/Walker2d/HalfCheetah/Ant)natural return: Hopper 3081 -> 2101; Halfcheetah 5048 -> 1412Hurt
train with ARPO instead of SPOPPO · MuJoCoworst-case return under attack: Ant -3107 -> 1648Improved
use bilevel KL-surrogate adversary (BARPO) without SPO guidance instead of ARPO (ablation)ARPO · MuJoConatural and worst-case return: natural +75%, +104%, +156%, +113%; worst +4%, +173%, +209%, +61%Improved
add SPO guidance to BARPO (ablation)BARPO · MuJoConatural return vs worst-case robustnessMixed
use BARPO instead of ARPO with SPO guidance (ablation)ARPO (w g) · MuJoCorobust return: +19%, +193%, +138% robust return on Hopper, Halfcheetah, AntImproved
use BAR-PPO instead of WocaR-PPOPPO · MuJoCo Walker2dworst-case return: 1064 -> 2699Improved
use BAR-PPO instead of SA-PPOPPO · MuJoCo Hopperworst-case robustness: -0.634 vs -0.636No 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