Decoding Quantum Errors with Directional Insights
![The relationship between phase lifetime ([latex] PLP_L [/latex]) and anisotropy strength (β) is explored for the [latex] \llbracket 162,2,9\rrbracket [/latex] Toric code, demonstrating how these parameters influence the system’s stability.](https://arxiv.org/html/2601.07240v1/x2.png)
A new approach to quantum error correction utilizes directional priors to improve decoding performance without modifying code structure.
Science
Graphing Security: How Network Structure Impacts Data Protection
![A secure storage scheme distributes three files across eight servers in such a way that designated pairs of servers can reliably retrieve specific files, achieving a source key capacity of [latex]\frac{1}{2}[/latex] as theoretically established and demonstrably constructed-a configuration pushing the boundaries of data redundancy and access control.](https://arxiv.org/html/2601.07340v1/fig0.png)
New research explores the fundamental limits of secure data storage across interconnected networks, revealing how graph topology affects the achievable key rates for protecting sensitive information.
Can AI Learn to Spot Contract Flaws?
New research explores how large language models can be trained to identify critical security vulnerabilities in smart contracts, even with limited data.
Codes and Compositions: A Hidden Connection
A new mathematical bridge links the structure of error-correcting codes to the combinatorics of integer partitions, offering fresh insights into code design.
Divide and Conquer: Coding for Efficient Multi-Receiver Communication
A new approach to function-correcting codes leverages domain partitioning to reduce redundancy and improve communication efficiency for multiple recipients.
Sync Your Security: Architecting Multi-Device FIDO2 Credentials
A new review explores how to securely synchronize FIDO2 credentials across multiple devices without compromising hardware-based security.
Quantization’s Challenge: Estimating Means in a Noisy World
New research tackles the problem of accurately estimating the mean of data when information is severely limited by quantization and complicated by adversarial attacks.
Squeezing More Performance from Large Language Models
A new quantization technique significantly improves the efficiency of running massive AI models on NVIDIA hardware.
Decoding Deception: Protecting Users from QR Code Phishing
Researchers have developed a new method to detect malicious intent hidden within visually complex QR codes, offering a critical defense against a growing wave of ‘quishing’ attacks.
Beyond LSTMs: Smarter Smart Contract Security with Transformers
A new approach to detecting vulnerabilities in Ethereum smart contracts leverages the power of Transformer networks to improve accuracy and generalization.