As artificial intelligence becomes increasingly embedded in digital services, ensuring that these systems are efficient and sustainable is becoming a key priority. A new research work titled “Promoting Sustainable Web Agents: Benchmarking and Estimating Energy Consumption through Empirical and Theoretical Analysis” addresses this challenge by studying the energy consumption of web-based AI agents and proposing methods to understand better and optimize their behavior.
The SustainML platform continues to evolve with new capabilities designed to simplify and improve decision-making in sustainable artificial intelligence workflows. In this update, SustainML introduces a new feature that enables users to compare multiple Hugging Face models directly within the platform, combining structured metadata with AI-generated insights.
The SustainML platform continues to evolve with new capabilities designed to simplify the development of sustainable artificial intelligence systems. In the latest update, SustainML introduces a new integration with the Hugging Face ecosystem that enables users to discover, inspect, and evaluate machine learning models directly within the platform.
A new research introduces SPARC, a framework designed to address this challenge by decoupling visual perception and reasoning into two distinct computational stages. By separating these processes, the approach enables more efficient use of computational resources while preserving, and in many cases improving, model performance.
More details on the methodology and experimental evaluation are provided in the paper “SPARC: Separating Perception And Reasoning Circuits for Test-time Scaling of Vision-Language Models” (arXiv:2602.06566).
SustainML continues to evolve to support increasingly complex and distributed AI optimization scenarios. As part of this evolution, we have integrated Remote Procedure Call (RPC) functionality using Fast DDS Pro to enhance control-plane interactions inside the framework.
SustainML is among these nine innovative projects dedicated to creating a sustainable ML framework for Green AI.
Plaza de la Encina 10-11, Núcleo 4, 2ª Pl.
28760 Tres cantos - Madrid (España)
This project has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No 101070408.
