{"id":"arxiv/2603.17681","url":"https://arxiv.org/abs/2603.17681","tags":["math.NT","stat.ML"],"score":0,"title":"Murmurations, Mestre--Nagao sums, and Convolutional Neural Networks for elliptic curves","domain":"math.NT","source":"arxiv","authors":["Joanna Bieri","Edgar Costa","Alyson Deines","Kyu-Hwan Lee","David Lowry-Duda","Thomas Oliver","Yidi Qi","Tamara Veenstra"],"reviews":0,"subject":"math.NT","summary":"We apply one-dimensional convolutional neural networks to the Frobenius traces of elliptic curves over $\\mathbb{Q}$ and evaluate and interpret their predictive capacity. In keeping with similar experiments by Kazalicki--Vlah, Bujanović--Kazalicki--Novak, and Pozdnyakov, we observe high accuracy predictions for the analytic rank across a range of conductors. We interpret the prediction using saliency curves and explore the interesting interplay between saliency, murmurations and Mestre--Nagao sums.","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-18","reactions":{}}