SPARC: A Multi-Agent System for Electrical Circuit Question Answering
On the map: LLM reasoning → Reasoning for agents, planning and code
also in 1 more
Abstract (arXiv)
Electrical circuit diagram QA tasks require complex mathematical reasoning, which remains challenging for multimodal LLMs. We present SPARC, a multi-agent system that answers questions over circuit diagrams by grounding reasoning in executable physics-based simulations. SPARC uses LLM agents to synthesize, execute, and analyze simulation programs, improving accuracy and reliability by design. It achieves 83% accuracy, with up to a 58% absolute improvement over baselines, while enabling systematic error diagnosis.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 11 of its 25 references that are in the map. First posted on arXiv 5 Jun 2026.
Builds on
11 of 25 references.
References · 11
- 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
- 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,516
- 2019OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledgecited by 1,121
- 2024MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmarkcited by 512
- 2019PlotQA: Reasoning over Scientific Plotscited by 321
- 2025jina-embeddings-v4: Universal Embeddings for Multimodal Multilingual Retrievalcited by 76
- 2024MATHSENSEI: A Tool-Augmented Large Language Model for Mathematical Reasoningcited by 34
- 2024EEE-Bench: A Comprehensive Multimodal Electrical And Electronics Engineering Benchmarkcited by 15
- 2025ToolVQA: A Dataset for Multi-step Reasoning VQA with External Toolscited by 13
- 2025AMSnet 2.0: A Large AMS Database with AI Segmentation for Net Detectioncited by 9
- 2025Follow the Flow: Fine-grained Flowchart Attribution with Neurosymbolic Agentscited by 4
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026CircuitLM: A Multi-Agent LLM-Aided Design Framework for Generating Circuit Schematics from Natural Language Prompts
- 2026CircuTutor: Transforming Static Circuit Problems into Intelligent and Dynamic Tutoring
- 2025AITEE -- Agentic Tutor for Electrical Engineering
- 2026LLM-as-an-Investigator: Evidence-First Reasoning for Robust Interactive Problem Diagnosis
- 2026CircuitReason-1k: Benchmarking Long-Horizon Visual-to-Symbolic Reasoning inElectrical Circuits
- 2025CIRCUIT: A Benchmark for Circuit Interpretation and Reasoning Capabilities of LLMs
- 2026RF-Agent: A Practical Framework for Building Language Agents for RFIC Design
- 2026AaLLM: An End-to-End Analog Circuit Design Framework from Topology Generation to Sizing Using Large Language Models