Controlled Agentic Planning&Reasoning for Mechanism Synthesis

2025-05-23

AI Summary

This work addresses planar mechanism synthesis by proposing a dual-agent large language model (LLM) framework that enables synergistic reasoning between natural language understanding and symbolic computation. Methodologically, it establishes a language–symbol two-layer closed-loop inference architecture: natural language specifications drive geometric configuration generation, dynamic modeling, and executable simulation code synthesis; symbolic regression, abstract property equation mapping, and feedback-anchor–guided iterative refinement—optimized via inter-layer distance functions—ensure robust convergence. Key contributions include: (1) the first language–symbol joint reasoning paradigm for mechanism synthesis; (2) the release of the MSynth benchmark and end-to-end synthesis validation; (3) empirical identification of LLM scale as a critical threshold for activating symbolic regression insight; and (4) ablation studies confirming the indispensability of each module. The framework demonstrates strong effectiveness and convergence in generating both geometric configurations and dynamic models.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Planning, Routing, and Scheduling: Planning with Language Models
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

This work presents a dual-agent Large Language Model (LLM)-based reasoning method for mechanism synthesis, capable of reasoning at both linguistic and symbolic levels to generate geometrical and dynamic outcomes. The model consists of a composition of well-defined functions that, starting from a natural language specification, references abstract properties through supporting equations, generates and parametrizes simulation code, and elicits feedback anchor points using symbolic regression and distance functions. This process closes an actionable refinement loop at the linguistic and symbolic layers. The approach is shown to be both effective and convergent in the context of planar mechanisms. Additionally, we introduce MSynth, a novel benchmark for planar mechanism synthesis, and perform a comprehensive analysis of the impact of the model components. We further demonstrate that symbolic regression prompts unlock mechanistic insights only when applied to sufficiently large architectures.

Problem

• Dual-agent LLM method for mechanism synthesis
• Generates geometric and dynamic outcomes via reasoning
• Introduces MSynth benchmark for planar mechanisms

Innovation

• Dual-agent LLM for linguistic and symbolic reasoning
• Generates simulation code with symbolic regression
• Closed-loop refinement with feedback anchor points