cid f49d989e904e799d06c75c6b268636b91017022f23ebf8048354f15a738c8bf4 · original · open access · via openalex
Title: GAPMA: Graph-Attention Cell Usage Prediction with Polarity-Minimized NPN Aggregation for Design-Specific Cell Composition Authors: Chung‐Kuan Cheng; Junyeong Jang; Andrew B. Kahng; B. H. Kang; Jakang Lee Venue: Zenodo (CERN European Organization for Nuclear Research) Year: 2026 Canonical: doi:10.5281/zenodo.21970035 URL: https://doi.org/10.5281/zenodo.21970035 Collector: openalex OA: true ABSTRACT Artifact for MLCAD 2026 paper 179, GAPMA: Graph-Attention Cell Usage Prediction with Polarity-Minimized NPN Aggregation for Design-Specific Cell Composition. This release packages the GAPMA implementation source and shipped results required to reproduce the artifact. The archive contains .gitmodules, LICENSE, README.md, requirements.txt, examples/, flow/, rtl/, tools/, the pretrained GAT checkpoint, the 183-cell feature vocabulary, and the P-to-NP aggregation table.