Category Physics

Zuchongzhi-3, a superconducting quantum computing prototype with 105 qubits and 182 couplers, has made significant advancements in random quantum circuit sampling. This prototype was successfully developed by a research team from the University of Science and Technology of China (USTC).

This prototype operates at a speed that is 1015 times faster than the fastest supercomputer currently available and one million times faster than the latest results published by Google. This achievement marks a milestone in enhancing the performance of quantum computation, following the success of Zuchongzhi-2...

A team of physicists at the SLAC National Accelerator Laboratory, in Menlo Park, California, generated the highest-current, highest-peak-power electron beams ever produced. The team has published their paper in Physical Review Letters.

For many years, scientists have been finding new uses for high-powered laser light, from splitting atoms to mimicking conditions inside other planets. For this new study, the research team upped the power of electron beams, giving them some of the same capabilities.

The idea behind the newer, more powerful beams was pretty simple, the team acknowledges; it was figuring out how to make it happen that was difficult. The basic idea is to pack as much charge as possible into the shortest amount of time...

Quantum computers could be made with fewer overall components, thanks to technology inspired by Schrödinger’s cat. A team of researchers from Amazon Web Services has used “bosonic cat qubits,” to improve the ability of quantum computers to correct errors. The demonstration of quantum error correction requiring reduced hardware overheads is reported in a paper published in Nature.

The system uses so-called cat qubits (qubits are the quantum equivalent to classical computing bits), which are designed to be resistant against certain types of noise and errors that might disrupt the output of quantum systems. This approach requires fewer overall components to achieve quantum error correction than other designs.

Quantum computers are prone to errors, ...

A team of Carnegie Mellon University researchers set out to see how accurately large language models (LLMs) can match the style of text written by humans. Their findings were recently published in the Proceedings of the National Academy of Sciences.

“We humans, we adapt how we write and how we speak to the situation. Sometimes we’re formal or informal, or there are different styles for different contexts,” said Alex Reinhart, lead author and associate teaching professor in the Department of Statistics & Data Science.

“What we learned is that LLMs, like ChatGPT and Llama, write a certain way, and they don’t necessarily adapt to the writing style...