Deep neural networks are at the heart of artificial intelligence, ranging from pattern recognition to large language and reasoning models like ChatGPT. The principle: during a training phase, the parameters of the network’s artificial neurons are optimized in such a way that they can carry out specific tasks, such as autonomously discovering objects or characteristic features in images.Deep neural networks are at the heart of artificial intelligence, ranging from pattern recognition to large language and reasoning models like ChatGPT. The principle: during a training phase, the parameters of the network’s artificial neurons are optimized in such a way that they can carry out specific tasks, such as autonomously discovering objects or characteristic features in images.[#item_full_content]

Over the past years, computer scientists have introduced increasingly sophisticated generative AI models that can produce personalized content following specific inputs or instructions. While image generation models are now widely used, many of them are unpredictable and precisely controlling the images they create remains a challenge.Over the past years, computer scientists have introduced increasingly sophisticated generative AI models that can produce personalized content following specific inputs or instructions. While image generation models are now widely used, many of them are unpredictable and precisely controlling the images they create remains a challenge.[#item_full_content]

Forecasting electricity demand in buildings is now more accurate with Group Encoding (GE), a new method that uses only existing device operation data. Developed by researchers at the Institute of Science Tokyo, the method improved prediction accuracy by 74% in real-world tests.Forecasting electricity demand in buildings is now more accurate with Group Encoding (GE), a new method that uses only existing device operation data. Developed by researchers at the Institute of Science Tokyo, the method improved prediction accuracy by 74% in real-world tests.[#item_full_content]

Detecting writing via artificial intelligence is a tricky dance: Doing it right means being effective at identifying it while being careful not to falsely accuse a human of employing it. And few tools strike the right balance.Detecting writing via artificial intelligence is a tricky dance: Doing it right means being effective at identifying it while being careful not to falsely accuse a human of employing it. And few tools strike the right balance.[#item_full_content]

If you find pop-up ads annoying, imagine how distracting they are for a person with heightened sensitivity.If you find pop-up ads annoying, imagine how distracting they are for a person with heightened sensitivity.[#item_full_content]

This summer, EPFL and ETH Zurich will release a large language model (LLM) developed on public infrastructure. Trained on the Alps supercomputer at the Swiss National Supercomputing Center (CSCS), the new LLM marks a milestone in open-source AI and multilingual excellence.This summer, EPFL and ETH Zurich will release a large language model (LLM) developed on public infrastructure. Trained on the Alps supercomputer at the Swiss National Supercomputing Center (CSCS), the new LLM marks a milestone in open-source AI and multilingual excellence.[#item_full_content]

ChatGPT works by analyzing vast amounts of text, identifying patterns and synthesizing them to generate responses to users’ prompts. Color metaphors like “feeling blue” and “seeing red” are commonplace throughout the English language, and therefore comprise part of the dataset on which ChatGPT is trained.ChatGPT works by analyzing vast amounts of text, identifying patterns and synthesizing them to generate responses to users’ prompts. Color metaphors like “feeling blue” and “seeing red” are commonplace throughout the English language, and therefore comprise part of the dataset on which ChatGPT is trained.[#item_full_content]

Quantum computers have operated under a significant limitation: They can run only one program at a time. These million-dollar machines demand exclusive use even for the smallest tasks, leaving much of their expensive and fast-running hardware idle and forcing researchers to endure long lines.Quantum computers have operated under a significant limitation: They can run only one program at a time. These million-dollar machines demand exclusive use even for the smallest tasks, leaving much of their expensive and fast-running hardware idle and forcing researchers to endure long lines.[#item_full_content]

Maxim Van de Wynckel, a researcher at the WISE lab of the Vrije Universiteit Brussel (VUB), successfully defended his Ph.D. in computer science on 30 June 2025. Over six years, he explored a major challenge facing many modern technologies: How can indoor positioning systems—which determine the location of people or objects inside buildings—work better together?Maxim Van de Wynckel, a researcher at the WISE lab of the Vrije Universiteit Brussel (VUB), successfully defended his Ph.D. in computer science on 30 June 2025. Over six years, he explored a major challenge facing many modern technologies: How can indoor positioning systems—which determine the location of people or objects inside buildings—work better together?[#item_full_content]

For all their impressive capabilities, large language models (LLMs) often fall short when given challenging new tasks that require complex reasoning skills.For all their impressive capabilities, large language models (LLMs) often fall short when given challenging new tasks that require complex reasoning skills.[#item_full_content]