2605.27209v1 May 26, 2026 cs.AI

Learning to Act under Noise: Enhancing Agent Robustness via Noisy Environments

Yaorui Shi
Yaorui Shi
Citations: 300
h-index: 9
Yuxin Chen
Yuxin Chen
Citations: 400
h-index: 8
Tat-Seng Chua
Tat-Seng Chua
Citations: 3,271
h-index: 31
Xiaodong Cai
Xiaodong Cai
Citations: 80
h-index: 4
Yu Wang
Yu Wang
Citations: 58
h-index: 2
Y. Zhang
Y. Zhang
Citations: 169
h-index: 7
Zhuowen Han
Zhuowen Han
Citations: 27
h-index: 3
Xunliang Cai
Xunliang Cai
Citations: 231
h-index: 8
Qi Gu
Qi Gu
Citations: 51
h-index: 5
Junfeng Fang
Junfeng Fang
Citations: 4
h-index: 1
An Zhang
An Zhang
Citations: 50
h-index: 4
Xiang Wang
Xiang Wang
Citations: 77
h-index: 2

Recent advances in large language models (LLMs) have facilitated the widespread deployment of LLMs as interactive agents capable of reasoning, planning, and tool use. Despite strong performance on existing benchmarks, such agents often exhibit notable degradation when deployed in real-world settings, where environments are inherently stochastic and imperfect. We argue that this discrepancy arises from a fundamental mismatch between idealized training settings and real-world interaction dynamics, where current paradigms rely on carefully curated task instructions and stable, well-controlled environments. To address this gap, we propose NoisyAgent, an agentic training framework that explicitly incorporates environmental imperfections into the agent learning process. We identify two major sources of interaction noise in real-world scenarios: user noise, which captures ambiguity and variability in user interaction, and tool noise, which reflects failures and anomalies in tool execution. We introduce such perturbations into the training pipeline by modifying user interaction patterns and simulating tool execution results within the training environment. To stabilize training while encouraging agents to handle increasingly challenging imperfections, noise is applied to only a subset of rollouts and progressively increased in difficulty as the model adapts to the current noise level. Extensive experiments demonstrate that our approach consistently improves agent robustness under noisy and dynamic environments. Our analysis reveals that training under noise conditions also yields performance gains on idealized benchmarks, suggesting that controlled exposure to environmental noise promotes more generalizable reasoning and decision-making behaviors. Our findings highlight the importance of modeling interaction imperfections for bridging the gap between agent training and real-world deployment.

0 Citations
0 Influential
15.5 Altmetric
77.5 Score
Original PDF

No Analysis Report Yet

This paper hasn't been analyzed by Gemini yet.

Log in to request an AI analysis.

댓글

댓글을 작성하려면 로그인하세요.

아직 댓글이 없습니다. 첫 번째 댓글을 남겨보세요!