Reproduce “Inspector: Conversational and Lightweight Analyzer of Analog Circuit Layouts Using LLM and CNNs” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Abril Cano Castro, Giuseppe Chiari, Michele Piccoli, Federico Viola, Davide Zoni
Inspector addresses semantic questions about analog-chip layouts by combining a language model with task-specific visual detectors. It converts GDSII layouts to PNG images, routes task-specific prompts to a detector, and uses detected components to produce natural-language answers. On tested SkyWater 130 nm layouts, Task C accuracy is 99%, versus 18% for InternLM-4KHD; accuracy combines metrics without a specified formula.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.