news

New miniature laboratories are ensuring that artificial intelligence (AI) doesn't make mistakes. They provide a controlled test environment where algorithms and AI models can be checked before being put to work under real-life conditions. The aim is for AI to work reliably. Anyone who develops an AI solution sometimes goes on a journey into the unknown.

Scientists have developed a geometric deep learning method that can create a coherent picture of neuronal population activity during cognitive and motor tasks across experimental subjects and conditions. In the parable of the blind men and the elephant, several blind men each describe a different part of an elephant they are touching - a sharp tusk, a flexible trunk, or a broad leg - and disagree about the animal's true nature.

The reproduction number R is often used as an indicator to predict how quickly an infectious disease will spread. researchers have developed a mathematical model that is just as easy to use but enables more accurate predictions than R. Their model is based on a reproduction matrix that takes into account the heterogeneity of society.

Engineers at EPFL and ETH Zurich have come up with a new approach for reducing traffic congestion in large cities during rush hour. Their system is designed to be both fair and effective. Rush-hour traffic congestion is a major headache for commuters and a real problem for city planners. Worldwide, a huge amount of time is wasted stuck in traffic.

The way we memorise information - a mathematical problem statement, for example - reveals the way we process it. A team from the University of Geneva , in collaboration with CY Cergy Paris University (CYU) and Bourgogne University (uB), has shown how different solving methods can alter the way information is memorised and even create false memories.

Researchers have published a programmable framework that overcomes a key computational bottleneck of optics-based artificial intelligence systems. In a series of image classification experiments, they used scattered light from a low-power laser to perform accurate, scalable computations using a fraction of the energy of electronics.

Boys are good at math, girls not so much? A study from the University of Zurich has analyzed the social mechanisms that contribute to the gender gap in math confidence. While peer comparisons seem to play a crucial role for boys, girls' subjective evaluations are more likely to be based on objective performance.

Researchers at EPFL use neural networks to study proprioception, the sense the brain uses to "know" the body's movement and position. How does your brain know the position and movement of your different body parts? The sense is known as proprioception, and it is something like a "sixth sense", allowing us to move freely without constantly watching our limbs.

Scientists show that our mental representations of mathematical problems influence our strategies for solving them. Solving arithmetic problems, even simple subtractions, involves mental representations whose influence remains to be clarified. Visualizing these representations would enable us to better understand our reasoning and adapt our teaching methods.

In theory, quantum computers vastly outperform classical computers in terms of computing speed. For them to do so in practice, it is necessary to design more and novel high-speed algorithms, says ETH supercomputing specialist Torsten Hoefler. Quantum computers promise to be capable of solving some computational problems much faster than classical computers.

Scientists have known that when a mouse embryo is developing, the cells that will become its spine and muscles switch specific genes on and off repeatedly, in a synchronous fashion. However, there are deep mysteries about how these cells synchronize. FMI researchers have now developed a mathematical model that not only better explains how spontaneous synchronization arises in a developing mouse embryo, but may also offer some fundamental clues about how other biological systems sync up.

It doesn't have to be a hurricane or a tsunami - even a simple running water tap induces a shock wave upon impact with the sink. Now, with the help of CSCS's supercomputer "Piz Daint", mathematician Siddhartha Mishra of ETH Zürich is working to overcome current barriers to simulating and comprehending highly turbulent flows.

EPFL mathematicians, in collaboration with Purdue University, have settled a 30-year-old question about spheres and 4-dimensional spaces. The results bring new light to the "Euler Class," one of the most powerful tools to understand complicated spaces. For mathematicians, "Euler Class" is one of the most powerful tools for understanding complicated spaces by cutting them into simpler pieces.

Neural networks are learning algorithms that approximate the solution to a task by training with available data. However, it is usually unclear how exactly they accomplish this. Two young Basel physicists have now derived mathematical expressions that allow one to calculate the optimal solution without training a network.

Metabolism is essential to all living organisms, and modeling the chemical reactions that sustain life is no easy task. Now, scientists have released REKINDLE, paving the way for more efficient and accurate modeling of metabolic processes thanks to deep-learning. The way an organism metabolizes nutrients is a complex process.

A method known as quantum key distribution has long held the promise of communication security not possible in conventional cryptography. For the first time, an international team of scientists, including researchers from EPFL, has demonstrated experimentally an approach to quantum key distribution based on high-quality quantum entanglement - offering much broader security guarantees than previous schemes.

At its heart, is mathematics an aesthetic discipline? Or what does it mean if someone finds a proof -beautiful- And what does mathematical beauty say about physical connections? To this day, mathematics includes beautiful facts that are also familiar to us in everyday life. The golden ratio, for example, has been revered since an­tiquity as representing the most aesthetically pleasing proportions.

How does a cell balance risk and speed when dividing? scientists have developed and experimentally tested the first mathematical theory that describes the cell's best strategy for dividing safely and efficiently. Cells go through a life cycle that includes growing to the right size, being equipped to perform its functions, and finally dividing into two new cells.

When and where a gene is transcribed in a living organism often depends on its physical interactions with distal genomic regulatory regions called enhancers. Researchers in the group of Luca Giorgetti have thrown light on how such interactions control transcription thanks to a novel ingenious experimental approach combined with mathematical modelling.

Researchers from EPFL have found a way to use only mathematics to automatically draw neurons in 3D, meaning we are getting closer to being able to build digital twins of brains. Santiago Ramón y Cajal, a Spanish physician from the turn of the 19 th century, is considered by most to be the father of modern neuroscience.