VLM-CAD: VLM-Optimized Collaborative Agent Design Workflow for Analog Circuit Sizing

TL;DR

This paper introduces VLM-CAD, a novel workflow that combines vision language models with structural parsing and Bayesian optimization to improve analog circuit sizing through robust, explainable multimodal reasoning.

Contribution

The paper presents VLM-CAD, integrating a neuro-symbolic parser and an explainable Bayesian optimizer to enhance reasoning accuracy and reliability in engineering tasks.

Findings

Significantly improves spatial reasoning accuracy in circuit design.

Maintains physics-based explainability of decisions.

Achieves low power consumption with runtime under 66 minutes.

Abstract

Vision Language Models (VLMs) have demonstrated remarkable potential in multimodal reasoning, yet they inherently suffer from spatial blindness and logical hallucinations when interpreting densely structured engineering content, such as analog circuit schematics. To address these challenges, we propose a Vision Language Model-Optimized Collaborative Agent Design Workflow for Analog Circuit Sizing (VLM-CAD) designed for robust, step-by-step reasoning over multimodal evidence. VLM-CAD bridges the modality gap by integrating a neuro-symbolic structural parsing module, Image2Net, which transforms raw pixels into explicit topological graphs and structured JSON representations to anchor VLM interpretation in deterministic facts. To ensure the reliability required for engineering decisions, we further propose ExTuRBO, an Explainable Trust Region Bayesian Optimization method. ExTuRBO serves as an explainable grounding engine, employing agent-generated semantic seeds to warm-start local searches and utilizing Automatic Relevance Determination to provide quantified evidence for the VLM's decisions. Experimental results on two complex circuit benchmarks demonstrate that VLM-CAD significantly enhances spatial reasoning accuracy and maintains physics-based explainability. VLM-CAD consistently satisfies complex specification requirements while achieving low power consumption, with a total runtime under 66 minutes, marking a significant step toward robust, explainable multimodal reasoning in specialized technical domains.