Skip to content

Admissable: Training Reinforcement Learning Agents against Adversarial Missingness

Sep 2026 · 0 citations · 23 references
Computer Science

TL;DR

An adversarial training algorithm is developed and shown to be effectiveness in increasing robustness against adversarial missingness on three MuJoCo benchmark environments and achieves better robustness on all three tasks.

Abstract

In order to make Reinforcement Learning algorithms applicable in real world scenarios, safety must be ensured even under adverse operating conditions. In this work, we consider the challenge of adversarial feature missingness: a scenario in which an adversary occludes features from the agent's observation in order to reduce performance as much as possible. We formally define adversarial missingness for Reinforcement Learning and compare it to the related concepts of $\ell_\infty$-norm bounded adversarial perturbations and learning with missing data. We develop an adversarial training algorithm and show its effectiveness in increasing robustness against adversarial missingness on three MuJoCo benchmark environments. Compared to a baseline trained with random uniform missingness, our method achieves better robustness on all three tasks.

View source

Similar papers

#machine learning Preprint Sep 2026

Robust Policy Optimization via Adversarial Importance Sampling

This work introduces Adversarial Importance Sampling (Advis), a method that uses importance sampling over trajectories from standard training to estimate and optimize verifiable worst-case returns and introduces advrl, a modular PyTorch library that provides clean, single-file implementations of existing robustness met...

Amine Andam, Jamal Bentahar, M. Hedabou · 0 citations
Conference Aug 2026

Robust Deep Reinforcement Learning via Adversarial Data Augmentation for Robotic Control Under Observation Uncertainty

Observation uncertainty can drive a deep reinforcement learning (DRL) controller to select inappropriate actions even when its nominal policy performs well. This issue is especially pronounced in robotic systems deployed outside controlled laboratory conditions, where sensing and communication are affected by noise, bi...

Hai-Tao Zhang, Cheng-Yu Liu, Qing-Kai Meng et al. · 0 citations
#machine learning Preprint Aug 2026

RL-FAT: Reinforcement Learning for Fair Adversarial Training

RL-FAT is proposed, a reinforcement-learning-inspired fair adversarial training framework that uses policy-gradient based feedback from adversarial predictions to improve adversarial robustness while promoting a more balanced robustness distribution across classes.

Tejaswini Medi, Levan Mikeladze, Margret Keuper · 0 citations
#machine learning Preprint Sep 2026

Robust Adversarial Reinforcement Learning with Risk Sensitivity and Critic Consistency Regularization

This work proposes a unified framework, RACER (Risk-sensitive robust Adversarial critic ConsistEncy-regularized Reinforcement learning), that revisits adversarial RL from a risk-sensitive perspective and introduces a state-dependent adversarial objective that adaptively regulates perturbation strength.

Jia-Xin Wu, Tian-Tian Zhang, Yu-Xing Wang et al. · 0 citations

Reinforcement Learning Inspired Black-box Adversarial Attacks for Computer Vision

This work proposes a novel approach under the more realistic black-box threat model that utilizes concepts from reinforcement learning to optimize perturbations with a non-differentiable target model, and shows the success of reinforcement learning inspired black-box adversarial attack (RIBA) in generating adversarial...

F. Krone, Elena Hoemann, Sven Hallerbach · 0 citations
#machine learning Preprint Oct 2026

Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures

Adversarial training is one of the most reliable defenses against adversarial attacks, but its high computational cost must generally be paid anew for each task. Robust foundation models offer a promising alternative: adversarially pretrain a model once and then transfer its robustness to downstream tasks through light...

Soichiro Kumano · 0 citations

Related blog posts

MIT News · Artificial Intelligence Nov 17, 2026

Learning the language of machine learning

By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.

Microsoft Research Blog Nov 16, 2026

TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient

AI coding agents are increasingly help developers fix bugs and build features largely on their own, but every interaction with their models consumes tokens. In each session, an agent sends instructions, tool descriptions, and session history to its model. Most efforts to reduce this cost focus on individual tasks. Yet developers return to the same […] The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.