Weak lensing machine learning challenge improves cosmology estimates

FAIR Universe Weak Lensing ML Uncertainty Challenge: Dataset and Cosmological Inference Competition

Artificial IntelligenceComputer Vision and Pattern RecognitionMachine Learning

Summary

Weak gravitational lensing lets scientists see how matter is spread across the universe, helping them learn about how the universe grows and changes. The authors created a special challenge with realistic data that includes common errors scientists face, so machine learning approaches can be tested fairly. They also collected the best methods from the challenge, which include new ways to handle data and make predictions about the cosmos. This work gives the community tools and data to better use machine learning for understanding the universe despite uncertainties.

What this means in practice

  • For cosmology data analysts: Use the challenge dataset and methods to benchmark and improve machine learning models for cosmological parameter estimation from weak lensing data.
  • For machine learning engineers: Develop and test robust ML architectures incorporating uncertainty handling for scientific datasets with systematic errors, using the provided benchmark.

Authors

Biwei Dai, Po-Wen Chang, Erwan Allys, Wahid Bhimji, Boris Bolliet, Thomas Borrett, Paolo Calafiura, Ragansu Chakkappai, Yuan-Tang Chou, Noé Dia, Sascha Diefenbacher, Ibrahim Elsharkawy, Steven Farrell, Sacha Guerrini, Isabelle Guyon, Chris Harris, Shih-Chieh Hsu, Wassim Kablan, François Lanusse, Celia Lecat, Julia Linhart, Yuezhou Ma, Benjamin Nachman, Shubhojit Naskar, Andy Nilipour, Laurence Perreault-Levasseur, Sébastien Pierre, Benjamin Remy, David Rousseau, Uroš Seljak, Haonan Shangguan, Sammy Nasser Sharief, Yuanxu Sun, Andreas Tersenov, Ihsan Ullah, Huikun Weng, Licong Xu, Justine Zeghal, Yulei Zhang, Hang Zhou

Abstract

Weak gravitational lensing is a powerful cosmological probe that enables direct mapping of the matter distribution and precise constraints on the fundamental parameters governing structure formation in the universe. In recent years, machine learning has shown strong potential for extracting cosmological information from weak-lensing data beyond traditional summary statistics. However, progress has been limited by the lack of standardized benchmarks that capture the complexity of the dataset, simulator imperfections, and various systematic biases encountered in realistic analyses. To address this gap, we present the FAIR Universe Weak Lensing Machine Learning Uncertainty Challenge, one of the NeurIPS 2025 competitions, as a benchmark dataset and evaluation study. Our dataset reflects realistic weak-lensing observations, incorporating multiple systematic biases, and thereby providing a controlled testbed for robust cosmological inference under uncertainties. We summarize the main methodological advances developed by the winners of our challenge, including data augmentation with conditional optimal transport, an AI-agent-driven approach for parameter inference, ensemble learning, Vision Transformer-based architectures, and physics-informed feature integration. By unifying the benchmark dataset and the winning solutions, this work provides a community resource for developing and evaluating machine learning methods for weak lensing cosmology and, more broadly, for scientific inference under systematic uncertainties.