Category Physics

When large language models (LLMs) make decisions about networking and friendship, the models tend to act like people, across both synthetic simulations and real-world network contexts.

Marios Papachristou and Yuan Yuan developed a framework to study network formation behaviors of multiple LLM agents and compared these behaviors against human behaviors. The paper is published in the journal PNAS Nexus.

How LLMs form network connections
The authors conducted simulations using several large language models placed in a network, which were asked to choose which other nodes to connect with, given their number of connections, common neighbors, and shared attributes, like arbitrarily assigned “hobbies” or “location.”

The authors varied the network context, including simulations of fri...

In physical systems, transport takes many forms, such as electric current through a wire, heat through metal, or even water through a pipe. Each of these flows can be described by how easily the underlying quantity—charge, energy, or mass—moves through a material.

Normally, collisions and friction lead to resistance causing these flows to slow down or fade away. But in a new experiment at TU Wien, scientists have observed a system where that doesn’t happen at all.

By confining thousands of rubidium atoms to move along a single line using magnetic and optical fields, they created an ultracold quantum gas in which energy and mass move with perfect efficiency...

Over the past decades, physicists and quantum engineers introduced a wide range of systems that perform desired functions leveraging quantum mechanical effects. These include so-called quantum sensors, devices that rely on qubits (i.e., units of quantum information) to detect weak magnetic or electric fields.

Researchers at the HUN-REN Wigner Research Center for Physics, the Beijing Computational Science Research Center, the University of Science and Technology of China and other institutes recently introduced a new quantum sensing platfor...

Quantum computers are computing systems that process information leveraging quantum mechanical effects. These computers rely on qubits (i.e., the quantum equivalent of bits), which can store information in a mixture of states, as opposed to binary states (0 or 1).

While quantum computers could tackle some computational and optimization problems faster and more effectively than classical computers, they are also inherently more prone to errors. This is because qubits can be easily disturbed by disturbances from their surrounding environment, also referred to as noise.

Over the past decades, quantum engineers and physicists have been trying to develop approaches to correct noise-related errors, also known as quantum error correction (QEC) techniques...