Paper Accepted and Presented in NeurIPS 2025

We are thrilled to announce that our collaborative research with University College London (UCL) was successfully accepted and presented at the Tackling Climate Change with Machine Learning workshop at NeurIPS 2025. The paper introduces a novel, domain-consistent neural network framework that integrates data-driven domains as constraints to solve nonlinear optimization challenges in gas power plants. By scaling our verified efficiency gains to the global fleet, we estimate a massive annual reduction of 26 million tons in CO₂ emissions, highlighting the scalable power of AI in driving near-term decarbonization. Please watch the embedded video presentation below to learn more about our methodology and findings.

Watch the presentation: https://www.climatechange.ai/papers/neurips2025/37