The Sim-to-Real Gap of Foundation Model Agents: A Unified MDP Perspective

arXiv:2606.07017v1 Announce Type: cross Abstract: Foundation model agents are increasingly deployed for real-world decision-making, but suffer from the sim-to-real gap. While robotics and classical control have mature frameworks to address this gap, the foundation model community is treating agent robustness as an entirely novel phenomenon. Our paper proposes formalizing the foundation model agent evaluation and training gap as a classical sim-to-real problem structured entirely around the four elements of a Markov Decision Process, including Observation, Action, Transition, and Reward. In this paper, we set a comprehensive research agenda that translates classical discrepancies into the foundation model domain and advocates for adopting established solutions like domain randomization. We provide concrete examples, such as a multilingual tool calling to demonstrate how severe observation space gaps lead to operationally invalid actions despite correct semantic intent. Ultimately, this agenda aims to drive a paradigm shift, yielding a unified vocabulary and standardized stress test benchmarks to foster a new generation of highly trustworthy agents for reliable real-world applications.

Abstract:Foundation model agents are increasingly deployed for real-world decision-making, but suffer from the sim-to-real gap. While robotics and classical control have mature frameworks to address this gap, the foundation model community is treating agent robustness as an entirely novel phenomenon. Our paper proposes formalizing the foundation model agent evaluation and training gap as a classical sim-to-real problem structured entirely around the four elements of a Markov Decision Process, including Observation, Action, Transition, and Reward. In this paper, we set a comprehensive research agenda that translates classical discrepancies into the foundation model domain and advocates for adopting established solutions like domain randomization. We provide concrete examples, such as a multilingual tool calling to demonstrate how severe observation space gaps lead to operationally invalid actions despite correct semantic intent. Ultimately, this agenda aims to drive a paradigm shift, yielding a unified vocabulary and standardized stress test benchmarks to foster a new generation of highly trustworthy agents for reliable real-world applications.

7 pages, 2 figures, 2 tables. Accepted by KDD 2026 Blue Sky Ideas Track

Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Emerging Technologies (cs.ET)

I.2.7; I.2.6; I.2.4

arXiv:2606.07017 [cs.AI]

(or arXiv:2606.07017v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2606.07017 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

https://doi.org/10.1145/3770855.3818660 Focus to learn more DOI(s) linking to related resources

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

analiza aiqbench

The Sim-to-Real Gap of Foundation Model Agents: A Unified MDP Perspective

Subiecte conexe