Capturing Quantum Correlations with Transformer Networks
![A theoretical model investigates itinerant spinful fermions coupled to ancillary spin layers-arranged as a ladder and interacting via Kondo exchange, Heisenberg interactions, and interlayer coupling-to create a composite local Hilbert space of dimension [latex]\mathcal{V} = 2^{4}[/latex], thereby exploring the interplay between fermionic and spin degrees of freedom within the system.](https://arxiv.org/html/2603.02316v1/2603.02316v1/x1.png)
A new approach uses Transformer neural networks to represent quantum states, offering improved accuracy in simulating complex many-body systems.
Science
Building Trust in Complex AI Systems
A new framework enables continuous verification of system integrity by composing cryptographic proofs for individual components.
Beyond Perturbation: Taming Resonances in Scattering Amplitudes
This review explores the essential techniques for extending the reach of effective field theories into regimes dominated by resonant behavior.
Beyond the Chain: Unmasking Blockchain’s Communication Weaknesses
A new analysis reveals critical vulnerabilities in how blockchains communicate, potentially undermining their security and resilience to attack.
The Privacy Price of Quantum Machine Learning
Combining quantum computing with federated learning promises enhanced privacy, but new research reveals significant resource demands when using fully homomorphic encryption.
Beyond Passphrases: Securing Wallet Recovery in a Decentralized World
A new protocol, VA-DAR, offers a robust and serverless solution for wallet recovery that prioritizes user privacy and resilience against account enumeration.
Silicon Qubits Weather the Noise Storm
![Engineered Zeeman splitting creates a linear energy gradient across a five-qubit array fabricated in [latex] ^{28}Si[/latex], enabling individual qubit addressability and positioning inference while simultaneously exposing the system to charge-noise-induced dephasing mediated by two-level fluctuators in surrounding oxides and a cobalt micromagnet, all within a device architecture designed for spin-to-charge conversion and controlled by global microwave rotations under an externally applied magnetic field.](https://arxiv.org/html/2603.03051v1/2603.03051v1/x1.png)
New research reveals that correlated noise in silicon spin qubits, while present, doesn’t preclude the path to scalable, fault-tolerant quantum computers.
Function-Correcting Codes: Limits to Enhanced Data Protection
New research reveals fundamental constraints on the ability of function-correcting codes to provide stronger error resilience than traditional data protection methods.
Securing the Loop: UAV-Based Systems Under Attack
This review explores how to protect sensitive data and maintain reliable control in unmanned aerial vehicle networks facing potential eavesdropping.
Shielding Federated Learning from Data Sleuths
New research offers a robust defense against attacks that aim to identify the origin of data used in collaborative machine learning.