Agent-Centric Visual Reinforcement Learning under Dynamic Perturbations

• URL: http://arxiv.org/abs/2604.24661v2
• Date: Tue, 28 Apr 2026 07:45:50 GMT
• Title: Agent-Centric Visual Reinforcement Learning under Dynamic Perturbations
• Abstract summary: We introduce the Visual Degraded Control Suite (VDCS) to simulate non-stationary real-world perturbations.<n>Experiments on VDCS reveal severe performance degradation in existing methods.<n>We propose Agent-Centric Observations with Mixture-of-Experts (ACO-MoE) to robustify visual RL against perturbations.
• Score: 24.021326345297535
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Visual reinforcement learning aims to empower an agent to learn policies from visual observations, yet it remains vulnerable to dynamic visual perturbations, such as unpredictable shifts in corruption types. To systematically study this, we introduce the Visual Degraded Control Suite (VDCS), a benchmark extending DeepMind Control Suite with Markov-switching degradations to simulate non-stationary real-world perturbations. Experiments on VDCS reveal severe performance degradation in existing methods. We theoretically prove via information-theoretic analysis that this failure stems from reconstruction-based objectives inevitably entangling perturbation artifacts into latent representations. To mitigate this negative impact, we propose Agent-Centric Observations with Mixture-of-Experts (ACO-MoE) to robustify visual RL against perturbations. The proposed framework leverages unique agent-centric restoration experts, achieving restoration from corruptions and task-relevant foreground extraction, thereby decoupling perception from perturbation before being processed by the RL agent. Extensive experiments on VDCS show our ACO-MoE outperforms strong baselines, recovering 95.3% of clean performance under challenging Markov-switching corruptions. Moreover, it achieves SOTA results on DMControl Generalization with random-color and video-background perturbations, demonstrating a high level of robustness.

Related papers

• ARISE-RL: Agentic Rubric-Grounded Iterative Self-Evolution with Reinforcement Learning [74.19730275669227]
  ARISE-RL is a novel full-cycle self-evolution framework for training open-ended agents.<n>We present Reward-Gated Self-Evolution Distillation (RG-SED), which distills a memory-augmented variant of the same policy back into itself.<n>We also present ECR-Bench, an expert-calibrated rubric benchmark suite covering single-tool deep research and multi-tool travel planning.
  arXiv  Detail & Related papers  (2026-09-01T10:54:13Z)
• IADD-TR: Intervention-Aware Dynamics Decoupling with Targeted Regularization for Model-Based Reinforcement Learning [57.16513298507272]
  We propose IADD-TR, a unified framework combining Intervention-Aware Dynamics Decoupling (IADD) and Targeted Regularization (TR)<n>IADD factorizes transitions into an action-intervention stage and an action-free natural evolution stage, using a zero-action anchor to resolve the non-uniqueness of this two-stage factorization for robust generalization.<n>For policy learning, we derive TR from the efficient influence function of a replay-state policy-gradient functional.
  arXiv  Detail & Related papers  (2026-08-11T08:20:02Z)
• HERO: Hindsight-Enhanced Reflection from Environment Observations for Agentic Self-Distillation [50.53459634301361]
  HERO is a hindsight-enhanced self-distillation framework that uses next environment observations as locally aligned feedback.<n> HERO improves task success and reduces unnecessary turns over environment-feedback-only self-distillation and GRPO.
  arXiv  Detail & Related papers  (2026-06-10T01:35:34Z)
• Mitigating Error Amplification in Fast Adversarial Training [58.74042726356826]
  Fast Adversarial Training (FAT) has proven effective in enhancing model robustness by encouraging networks to learn perturbation-invariant representations.<n>FAT often suffers from catastrophic overfitting (CO), where the model overfits to the training attack and fails to generalize to unseen ones.<n>We propose a Distribution-aware Dynamic Guidance (DDG) strategy that dynamically adjusts both the perturbation budget and supervision signal.
  arXiv  Detail & Related papers  (2026-04-27T11:23:18Z)
• Improving Robustness of Vision-Language-Action Models by Restoring Corrupted Visual Inputs [6.2827295422415235]
  Vision-Language-Action (VLA) models have emerged as a dominant paradigm for generalist robotic manipulation.<n>However, reliable real-world deployment is severely hindered by their fragility to visual disturbances.<n>We introduce the Corruption Restoration Transformer (CRT), a vision transformer designed to immunize VLA models against sensor disturbances.
  arXiv  Detail & Related papers  (2026-02-01T11:09:08Z)
• Lost in the Noise: How Reasoning Models Fail with Contextual Distractors [57.31788955167306]
  Recent advances in reasoning models and agentic AI systems have led to an increased reliance on diverse external information.<n>We introduce NoisyBench, a comprehensive benchmark that systematically evaluates model robustness across 11 datasets in RAG, reasoning, alignment, and tool-use tasks.<n>Our evaluation reveals a catastrophic performance drop of up to 80% in state-of-the-art models when faced with contextual distractors.
  arXiv  Detail & Related papers  (2026-01-12T05:43:51Z)
• Robust-R1: Degradation-Aware Reasoning for Robust Visual Understanding [54.05243949024302]
  Existing robust MLLMs rely on implicit training/adaptation that focuses solely on visual encoder generalization.<n>We propose Robust-R1, a novel framework that explicitly models visual degradations through structured reasoning chains.<n>Our approach integrates: (i) supervised fine-tuning for degradation-aware reasoning foundations, (ii) reward-driven alignment for accurately perceiving degradation parameters, and (iii) dynamic reasoning depth scaling adapted to degradation intensity.
  arXiv  Detail & Related papers  (2025-12-19T12:56:17Z)
• Deep Learning Based Dense Retrieval: A Comparative Study [11.705651144832041]
  We assess the vulnerability of dense retrieval systems to poisoned tokenizers by evaluating models such as BERT, Dense Passage Retrieval (DPR), Contriever, SimCSE, and ANCE. Our experiments reveal that even small perturbations can severely impact retrieval accuracy, highlighting the need for robust defenses in critical applications.
  arXiv  Detail & Related papers  (2024-10-27T02:52:36Z)
• DEAR: Disentangled Environment and Agent Representations for Reinforcement Learning without Reconstruction [4.813546138483559]
  Reinforcement Learning (RL) algorithms can learn robotic control tasks from visual observations, but they often require a large amount of data. In this paper, we explore how the agent's knowledge of its shape can improve the sample efficiency of visual RL methods. We propose a novel method, Disentangled Environment and Agent Representations, that uses the segmentation mask of the agent as supervision.
  arXiv  Detail & Related papers  (2024-06-30T09:15:21Z)
• Towards Evaluating the Robustness of Visual State Space Models [63.14954591606638]
  Vision State Space Models (VSSMs) have demonstrated remarkable performance in visual perception tasks. However, their robustness under natural and adversarial perturbations remains a critical concern. We present a comprehensive evaluation of VSSMs' robustness under various perturbation scenarios.
  arXiv  Detail & Related papers  (2024-06-13T17:59:44Z)
• VIBR: Learning View-Invariant Value Functions for Robust Visual Control [3.2307366446033945]
  VIBR (View-Invariant Bellman Residuals) is a method that combines multi-view training and invariant prediction to reduce out-of-distribution gap for RL based visuomotor control. We show that VIBR outperforms existing methods on complex visuo-motor control environment with high visual perturbation.
  arXiv  Detail & Related papers  (2023-06-14T14:37:34Z)
• Policy Smoothing for Provably Robust Reinforcement Learning [109.90239627115336]
  We study the provable robustness of reinforcement learning against norm-bounded adversarial perturbations of the inputs. We generate certificates that guarantee that the total reward obtained by the smoothed policy will not fall below a certain threshold under a norm-bounded adversarial of perturbation the input.
  arXiv  Detail & Related papers  (2021-06-21T21:42:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.