![[2506.04500] "Don't Do That!": Guiding Embodied Systems through Large Language Model-based Constraint Generation](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2506.04500] "Don't Do That!": Guiding Embodied Systems through Large Language Model-based Constraint Generation
Summary
This paper presents STPR, a framework that utilizes large language models to convert complex natural language constraints into executable Python functions for robotic navigation.
Why It Matters
As robotics increasingly incorporates natural language processing, this research addresses the challenge of translating informal constraints into formal programming. By leveraging LLMs, the framework enhances the efficiency and accuracy of robotic navigation, which is crucial for advancing AI capabilities in real-world applications.
Key Takeaways
- STPR framework translates natural language constraints into Python code.
- Utilizes LLMs to avoid complex reasoning and potential errors.
- Demonstrates effectiveness in simulated environments with various constraints.
- Compatible with smaller, code-specific LLMs for cost-effective implementation.
- Enhances robotic navigation by ensuring compliance with complex constraints.
Computer Science > Artificial Intelligence arXiv:2506.04500 (cs) [Submitted on 4 Jun 2025 (v1), last revised 24 Feb 2026 (this version, v2)] Title:"Don't Do That!": Guiding Embodied Systems through Large Language Model-based Constraint Generation Authors:Amin Seffo, Aladin Djuhera, Masataro Asai, Holger Boche View a PDF of the paper titled "Don't Do That!": Guiding Embodied Systems through Large Language Model-based Constraint Generation, by Amin Seffo and 3 other authors View PDF HTML (experimental) Abstract:Recent advancements in large language models (LLMs) have spurred interest in robotic navigation that incorporates complex spatial, mathematical, and conditional constraints from natural language into the planning problem. Such constraints can be informal yet highly complex, making it challenging to translate into a formal description that can be passed on to a planning algorithm. In this paper, we propose STPR, a constraint generation framework that uses LLMs to translate constraints (expressed as instructions on ``what not to do'') into executable Python functions. STPR leverages the LLM's strong coding capabilities to shift the problem description from language into structured and transparent code, thus circumventing complex reasoning and avoiding potential hallucinations. We show that these LLM-generated functions accurately describe even complex mathematical constraints, and apply them to point cloud representations with traditional search algorithms. Experiments ...