English

ReDAct: Uncertainty-Aware Deferral for LLM Agents

Computation and Language 2026-04-09 v1 Machine Learning Multiagent Systems

Abstract

Recently, LLM-based agents have become increasingly popular across many applications, including complex sequential decision-making problems. However, they inherit the tendency of LLMs to hallucinate, leading to incorrect decisions. In sequential settings, even a single mistake can irreversibly degrade the trajectory, making hallucinations an even bigger problem. Although larger LLMs hallucinate less, they incur a significantly higher per-token cost. In this paper, we address this tradeoff by proposing ReDAct (Reason-Defer-Act). In ReDAct, an agent is equipped with two LLMs: a small, cheap model used by default, and a large, more reliable but expensive model. When the predictive uncertainty of the small model exceeds a calibrated threshold, the decision is deferred to the large model. We evaluate our approach in text-based embodied environments such as ALFWorld and MiniGrid and show that deferring only about 15% of decisions to the large model can match the quality of using it exclusively, while significantly reducing inference costs.

Keywords

Cite

@article{arxiv.2604.07036,
  title  = {ReDAct: Uncertainty-Aware Deferral for LLM Agents},
  author = {Dzianis Piatrashyn and Nikita Kotelevskii and Kirill Grishchenkov and Nikita Glazkov and Ivan Nasonov and Ilya Makarov and Timothy Baldwin and Preslav Nakov and Roman Vashurin and Maxim Panov},
  journal= {arXiv preprint arXiv:2604.07036},
  year   = {2026}
}
R2 v1 2026-07-01T11:59:14.110Z