AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Robust Deep Reinforcement Learning against Adversarial Behavior Manipulation

Shojiro Yamabe, Kazuto Fukuchi, Jun Sakuma

ICLR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Attacks and defenses in sequential settings · 2 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
add time-discounting (gamma^t) to the smoothness regularizer (ablation)SA-PPO policy smoothing · Meta-World (10 tasks)clean reward: avg +28.2%; drawer-close 2156 -> 4237Improved
add time-discounting to policy smoothing (ablation)SA-PPO · Meta-World drawer-closebest attack reward: 4 -> 4860Hurt
add time-discounting to policy smoothing (ablation)SA-PPO · Meta-World window-closebest attack reward: 485 -> 482No clear effect
add time-discounting to policy smoothing (ablation)SA-PPO · Meta-World handle-pull-sidebest attack reward: 10 -> 7No clear effect
add time-discounting to policy smoothing (ablation)SA-PPO · Meta-World faucet-closebest attack reward: 1559 -> 1789Hurt
add policy smoothing (SA-PPO or TDRT) to PPOPPO · Meta-World window-closebest attack reward: 4505 -> 482Improved
adversarial training (ATLA-PPO, PA-ATLA-PPO) added to PPOPPO · Meta-World window-closebest attack reward: 4505 -> 4270 / 4041No clear effect
use imitation-learning attack BIA instead of targeted PGDadversary · Meta-World drawer-closeattack reward: 2891 -> 4760Improved
remove victim action access (ILfO no-box vs ILfD black-box)BIA · Meta-World faucet-openattack reward: 3031 -> 2718Hurt
BIA (demonstrations only) vs reward-maximization SA-RLBIA · Meta-World faucet-openattack reward: 1448 -> 3031Improved
BIA vs SA-RLBIA · Meta-World window-closeattack reward: 4505 -> 3962Hurt

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