AutoVSR Automatic Visual-to-Symbolic Reasoning for Symbolic Expression Generation from Circuit Schematic

Jul 16, 2026

Symbolic expressions enable accurate characterization and prediction of circuit behavior, yet deriving them directly from circuit schematics is difficult because it demands both reliable visual-to-symbolic circuit structure construction and correct multi-step symbolic derivation. AutoVSR presents an automated visual-to-symbolic generation framework using Vision Language Models (VLMs). It reconstructs schematics into an executable intermediate representation (Executable IR) and applies a symbolic solver for reasoning, improving accuracy by 30.01–59.45% and 41.96–51.84%, and outperforming closed-source VLMs in inference cost and efficiency.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages22

Reading Time55 min

Abstract

Introduction

Motivation for symbolic circuit analysis

Related work and research gap

Proposed AutoVSR overview

AutoVSR: Automatic Visual-to-Symbolic Reasoning for Symbolic Expression Generation from Circuit Schematic Zhe Xiao * 1 Longfei Li * 1 Xu He † 1 Haoying Wu 2 Zixing Zhang 1 Mingyu Liu 3 arXiv :2607 . 1 1338v 1 [ cs .AI] 13 Jul 2026 Abstract Symbolic expressions can effectively characterize and predict circuit behavior, but deriving them directly from circuit schematics is challenging. This process requires accurate visual-to-symbolic construction of circuit structure from images and correct multi-step symbolic derivation, both of which impose strict correctness requirements. This work proposes AutoVSR, an automated framework for visual-to-symbolic generation of circuit expressions using Vision Language Models (VLMs) . By reconstructing circuit diagrams into an executable intermediate representation (Executable IR) and leveraging a symbolic solver for reasoning, AutoVSR significantly improves the accuracy of symbolic expression generation. AutoVSR introduces two key innovations: an IR construction method guided by component rule retrieval and verification-based feedback, and a symbolic solver implemented as a planning agent equipped with a symbolic tool library for reliable multi-step derivation. Compared with end-to-end VLM approaches and specialized methods on the main symbolic expression generation task, AutoVSR achieves accuracy improvements of 30.01– 59.45% and 41.96–51.84%, respectively. Moreover, AutoVSR surpasses closed-source state-ofthe-art VLMs in inference cost and computational efficiency. Code is available at [https:](https:)//[github.com/LongfeiLi1/AutoVSR](github.com/LongfeiLi1/AutoVSR). 1. Introduction Integrated circuits form the physical foundation of the modern information society, enabling key functions such as information transmission, signal processing, and power *Equal contribution. † Xu He is the corresponding author, email: [dawn.hx@gmail.com](dawn.hx@gmail.com. 1Hunan University)[.](dawn.hx@gmail.com. 1Hunan University)[ 1](dawn.hx@gmail.com. 1Hunan University)[Hunan University](dawn.hx@gmail.com. 1Hunan University), Changsha, China 2Wuhan University of Technology, Wuhan, China 3Huazhong University of Science and Technology, Wuhan, China. Figure 1. AutoVSR Framework Workflow & Performance Results management in electronic systems (Hao et al., 2021) . To meet performance requirements such as stability and dynamic response, engineers often rely on symbolic analysis to characterize and predict circuit behavior during the design process (Gielen et al., 1994) . A key step in symbolic analysis is to convert the structural information in visual circuit schematics into computable symbolic expressions that describe circuit behavior, such as input–output relationships and related analytical forms including transfer functions (Gielen & Sansen, 1991) . In practice, obtaining these expressions requires interpreting schematic images to identify components and recover connectivity, followed by stepby-step symbolic derivation and simplification. However, this expertise-dependent workflow is labor-intensive. Therefore, an automated visual-to-symbolic reasoning framework is needed to bridge schematic images and symbolic expressions, improving efficiency and shortening design cycles. The emergence of multimodal large language models (MLLMs) has brought new opportunities to circuit-related tasks (Chen et al., 2024) . Current work mainly falls into two categories. (1) Circuit topology generation and reconstruction. These works focus on inferring the connectivity among circuit components, such as recovering netlists from schematics (Huang et al., 2025 ; Bhandari et al., 2025 ; Kulkarni et al., 2025) or synthesizing new circuit topologies (Vijayaraghavan et al., 2025 ; Gao et al., 2025) . (2) Assisted circuit sizing and optimization. Under a fixed topology, these works optimize device sizes and bias parameters (Yin et al., 2025 ; Karthik Somayaji & Li, 2025), typically by leveraging MLLMs to guide pa

FAQs

What is the main problem AutoVSR addresses in circuit symbolic expression generation?

AutoVSR targets the difficulty of generating correct symbolic expressions directly from circuit schematic images, which requires both accurate visual-to-symbolic construction of circuit structure and correct multi-step symbolic derivation.

How does AutoVSR improve reliability compared with end-to-end VLM approaches?

AutoVSR reconstructs schematics into an executable intermediate representation (Executable IR) and then uses a symbolic solver for reasoning, including verification-based feedback, rather than producing intermediate reasoning and final expressions purely end-to-end from a VLM.

What are the two key innovations introduced by AutoVSR?

It introduces (1) an IR construction method guided by component rule retrieval and verification-based feedback, and (2) a planning-agent symbolic solver equipped with a symbolic tool library to support reliable multi-step derivation.

Jul 16, 2026

Upload by Genevieve