{"ID":2867944,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.18384","arxiv_id":"2509.18384","title":"LAD-VF: LLM-Automatic Differentiation Enables Fine-Tuning-Free Robot Planning from Formal Methods Feedback","abstract":"Large language models (LLMs) can translate natural language instructions into executable action plans for robotics, autonomous driving, and other domains. Yet, deploying LLM-driven planning in the physical world demands strict adherence to safety and regulatory constraints, which current models often violate due to hallucination or weak alignment. Traditional data-driven alignment methods, such as Direct Preference Optimization (DPO), require costly human labeling, while recent formal-feedback approaches still depend on resource-intensive fine-tuning. In this paper, we propose LAD-VF, a fine-tuning-free framework that leverages formal verification feedback for automated prompt engineering. By introducing a formal-verification-informed text loss integrated with LLM-AutoDiff, LAD-VF iteratively refines prompts rather than model parameters. This yields three key benefits: (i) scalable adaptation without fine-tuning; (ii) compatibility with modular LLM architectures; and (iii) interpretable refinement via auditable prompts. Experiments in robot navigation and manipulation tasks demonstrate that LAD-VF substantially enhances specification compliance, improving success rates from 60% to over 90%. Our method thus presents a scalable and interpretable pathway toward trustworthy, formally-verified LLM-driven control systems.","short_abstract":"Large language models (LLMs) can translate natural language instructions into executable action plans for robotics, autonomous driving, and other domains. Yet, deploying LLM-driven planning in the physical world demands strict adherence to safety and regulatory constraints, which current models often violate due to hal...","url_abs":"https://arxiv.org/abs/2509.18384","url_pdf":"https://arxiv.org/pdf/2509.18384v2","authors":"[\"Yunhao Yang\",\"Junyuan Hong\",\"Gabriel Jacob Perin\",\"Zhiwen Fan\",\"Li Yin\",\"Zhangyang Wang\",\"Ufuk Topcu\"]","published":"2025-09-22T20:14:32Z","proceeding":"cs.RO","tasks":"[\"cs.RO\",\"cs.FL\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}