AnalogRetriever: Learning Cross-Modal Representations for Analog Circuit Retrieval

2026-04-25

AI Summary

This work addresses the long-standing limitation in analog circuit design where cross-modal semantic retrieval across heterogeneous modalities—SPICE netlists, schematics, and functional descriptions—has been constrained by single-modality exact matching. The paper proposes the first unified tri-modal retrieval framework that maps all three modalities into a shared embedding space via curriculum contrastive learning, enabling end-to-end cross-modal semantic alignment. Specifically, a vision-language model encodes schematics and textual descriptions, while a port-aware relational graph convolutional network is designed to process netlists. A high-quality repaired dataset is constructed to achieve 100% netlist usability. Evaluated on six cross-modal retrieval tasks, the method achieves an average Recall@1 of 75.2%, substantially outperforming baselines. Integration into AnalogCoder significantly improves functional pass rates and successfully solves previously infeasible design tasks.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Analogy

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Analog circuit design relies heavily on reusing existing intellectual property (IP), yet searching across heterogeneous representations such as SPICE netlists, schematics, and functional descriptions remains challenging. Existing methods are largely limited to exact matching within a single modality, failing to capture cross-modal semantic relationships. To bridge this gap, we present AnalogRetriever, a unified tri-modal retrieval framework for analog circuit search. We first build a high-quality dataset on top of Masala-CHAI through a two-stage repair pipeline that raises the netlist compile rate from 22\% to 100\%. Built on this foundation, AnalogRetriever encodes schematics and descriptions with a vision-language model and netlists with a port-aware relational graph convolutional network, mapping all three modalities into a shared embedding space via curriculum contrastive learning. Experiments show that AnalogRetriever achieves an average Recall@1 of 75.2\% across all six cross-modal retrieval directions, significantly outperforming existing baselines. When integrated into the AnalogCoder agentic framework as a retrieval-augmented generation module, it consistently improves functional pass rates and enables previously unsolved tasks to be completed. Our code and dataset will be released.

Problem

• analog circuit retrieval
• cross-modal representation
• SPICE netlist
• schematic
• functional description

Innovation

• cross-modal retrieval
• analog circuit design
• relational graph convolutional network
• vision-language model
• curriculum contrastive learning