I have liked kdnuggets for a while now. I used it for information on Tensorflow. This is cool though:

Neural networks are the building blocks behind every advanced AI system nowadays: from computer vision solutions to generative AI solutions and language models, most real-world solutions that involve some degree of AI have intricate neural network architectures at their core. But, what are neural networks and how do they perform surprisingly well in intelligently solving challenging tasks? To satisfy your curiosity at no cost, this post lists five resources to help you understand the mechanisms behind neural networks.

Here are five free resources in diverse formats and difficulty levels to acquaint with deep learning models at no cost.

China has reached a new milestone in quantum computing with the development of Tianyan-504, a powerful 504-qubit quantum computer.

The Tianyan-504 quantum computer was developed through collaboration between the China Telecom Quantum Group (CTQG), the Center for Excellence in Quantum Information and Quantum Physics under the Chinese Academy of Sciences, and QuantumCTek, a quantum technology company based in Anhui Province.

China has made a significant leap in quantum computing with the unveiling of the Tianyan-504, a record-breaking quantum computer.

Photosynthesis is one of the most efficient natural processes for converting light energy from the sun into chemical energy vital for life on earth. Proteins called photosystems are critical to this process and are responsible for the conversion of light energy to chemical energy.

Combining one kind of these proteins, called photosystem I (PSI), with platinum nanoparticles, microscopic particles that can perform a chemical reaction that produces hydrogen — a valuable clean energy source — creates a biohybrid catalyst. That is, the light absorbed by PSI drives hydrogen production by the platinum nanoparticle.

In a recent breakthrough, researchers at the U.S. Department of Energy’s (DOE) Argonne National Laboratory and Yale University have determined the structure of the PSI biohybrid solar fuel catalyst. Building on more than 13 years of research pioneered at Argonne, the team reports the first high-resolution view of a biohybrid structure, using an electron microscopy method called cryo-EM. With structural information in hand, this advancement opens the door for researchers to develop biohybrid solar fuel systems with improved performance, which would provide a sustainable alternative to traditional energy sources.

Argonne and Yale researchers shed light on the structure of a photosynthetic hybrid for the first time, enabling advancements in clean energy production.

Google DeepMind has announced an impressive grab bag of new products and prototypes that may just let it seize back its lead in the race to turn generative artificial intelligence into a mass-market concern.

Top billing goes to Gemini 2.0—the latest iteration of Google DeepMind’s family of multimodal large language models, now redesigned around the ability to control agents—and a new version of Project Astra, the experimental everything app that the company teased at Google I/O in May.

MIT Technology Review got to try out Astra in a closed-door live demo last week. It was a stunning experience, but there’s a gulf between polished promo and live demo.

Google just launched a ton of new products—including Gemini 2.0, which could power a new world of agents. And we got a first look.

To predict your #longevity, you have two main options. You can rely on the routine tests and measurements your doctor likes to order for you, such as blood pressure, cholesterol levels, weight, and so on. Or you can go down a biohacking rabbit hole the way tech millionaire turned longevity guru Bryan Johnson did to live longer. Johnson’s obsessive self-measurement protocol involves tracking more than a hundred biomarkers, ranging from the telomere length in blood cells to the speed of his urine stream (which, at 25 milliliters per second, he reports, is in the 90th percentile of 40-year-olds).

Scientists crunched the numbers to come up with the single best predictor of how long you’ll live—and arrived at a surprisingly low-tech answer.

One year of treatment with the targeted drug olaparib improves long-term survival in women with high-risk, early-stage breast cancer with mutations in BRCA1 or BRCA2 genes, new results from a major clinical trial show.

Ten years since the first patient was recruited, new findings from the phase III OlympiA trial – presented at San Antonio Breast Cancer Symposium (SABCS) 2024 – show that adding olaparib to standard treatment cuts the risk of cancer coming back by 35 per cent, and the risk of women dying by 28 per cent.

After six years, 87.5 per cent of patients who were treated with the drug were still alive compared with 83.2 per cent of those who were given the placebo pills.

Professor Andrew Tutt at The Institute of Cancer Research, London, and King’s College London is the global lead investigator and Chair of the Steering Committee for the OlympiA study, and was also involved in early laboratory research on PARP inhibitors such as olaparib, and their subsequent clinical development. The Breast International Group (BIG) coordinated the international OlympiA study, involving 671 study locations, globally across multiple partners. BIG coordinated the trial’s UK sites through the ICR Clinical Trials and Statistics Unit (ICR-CTSU).

When people say moons and Uranus it is phonetically a bit nasty, but, even but sounds a bit strange.

Instead, they may have oceans, and the moons may even be capable of supporting life, scientists say. Much of what we know about them was gathered by NASA’s Voyager 2 spacecraft which visited nearly 40 years ago.

S visit coincided with a powerful solar storm, which led to a misleading idea of what the Uranian system is really like. +Uranus is a beautiful, icy ringed world in the outer reaches of our solar system. It is among the coldest of all the planets. It is also tilted on its side compared to all the other worlds – as if it had been knocked over – making it arguably the weirdest.

The planet Uranus and its five biggest moons may not be the dead sterile worlds that scientists have long thought.

Instead, they may have oceans, and the moons may even be capable of supporting life, scientists say.

Much of what we know about them was gathered by NASA’s Voyager 2 spacecraft which visited nearly 40 years ago.

UNIVERSITY PARK, Pa. — A recently developed electronic tongue is capable of identifying differences in similar liquids, such as milk with varying water content; diverse products, including soda types and coffee blends; signs of spoilage in fruit juices; and instances of food safety concerns. The team, led by researchers at Penn State, also found that results were even more accurate when artificial intelligence (AI) used its own assessment parameters to interpret the data generated by the electronic tongue.

(Many people already posted this. This is the press release from Penn Sate who did the research)

The tongue comprises a graphene-based ion-sensitive field-effect transistor, or a conductive device that can detect chemical ions, linked to an artificial neural network, trained on various datasets. Critically, Das noted, the sensors are non-functionalized, meaning that one sensor can detect different types of chemicals, rather than having a specific sensor dedicated to each potential chemical. The researchers provided the neural network with 20 specific parameters to assess, all of which are related to how a sample liquid interacts with the sensor’s electrical properties. Based on these researcher-specified parameters, the AI could accurately detect samples — including watered-down milks, different types of sodas, blends of coffee and multiple fruit juices at several levels of freshness — and report on their content with greater than 80% accuracy in about a minute.

“After achieving a reasonable accuracy with human-selected parameters, we decided to let the neural network define its own figures of merit by providing it with the raw sensor data. We found that the neural network reached a near ideal inference accuracy of more than 95% when utilizing the machine-derived figures of merit rather than the ones provided by humans,” said co-author Andrew Pannone, a doctoral student in engineering science and mechanics advised by Das. “So, we used a method called Shapley additive explanations, which allows us to ask the neural network what it was thinking after it makes a decision.”

This approach uses game theory, a decision-making process that considers the choices of others to predict the outcome of a single participant, to assign values to the data under consideration. With these explanations, the researchers could reverse engineer an understanding of how the neural network weighed various components of the sample to make a final determination — giving the team a glimpse into the neural network’s decision-making process, which has remained largely opaque in the field of AI, according to the researchers. They found that, instead of simply assessing individual human-assigned parameters, the neural network considered the data it determined were most important together, with the Shapley additive explanations revealing how important the neural network considered each input data.

The semiconductor industry’s long held imperative—Moore’s Law, which dictates that transistor densities on a chip should double roughly every two years—is getting more and more difficult to maintain. The ability to shrink down transistors, and the interconnects between them, is hitting some basic physical limitations. In particular, when copper interconnects are scaled down, their resistivity skyrockets, which decreases how much information they can carry and increases their energy draw.

The industry has been looking for alternative interconnect materials to prolong the march of Moore’s Law a bit longer. Graphene is a very attractive optionin many ways: The sheet-thin carbon material offers excellent electrical and thermal conductivity, and is stronger than diamond.

However, researchers have struggled to incorporate graphene into mainstream computing applications for two main reasons. First, depositing graphene requires high temperatures that are incompatible with traditional CMOS manufacturing. And second, the charge carrier density of undoped, macroscopic graphene sheets is relatively low.

Making smaller transistors, and the interconnections between them, is getting near impossible. Copper interconnects get more resistive as they are scaled down, making them worse and slower at carrying information. Startup Destination 2D thinks graphene is the solution. They have a novel technique of growing graphene that is CMOS compatible, promising 100x current density improvement over copper.

Researchers have developed a device that can simultaneously measure six markers of brain health. The sensor, which is inserted through the skull into the brain, can pull off this feat thanks to an artificial intelligence (AI) system that pieces apart the six signals in real time.

Being able to continuously monitor biomarkers in patients with traumatic brain injury could improve outcomes by catching swelling or bleeding early enough for doctors to intervene. But most existing devices measure just one marker at a time. They also tend to be made with metal, so they can’t easily be used in combination with magnetic resonance imaging.

Simultaneous access to measurements could improve outcomes for brain injuries.