Patched, But Still Vulnerable?
![A residual risk score is computed by integrating semantic similarity [latex]S_{sem} [/latex], structural similarity [latex]S_{ast} [/latex], and cross-model agreement [latex] \sigma_{cross}^2 [/latex], offering a comprehensive assessment of potential vulnerabilities.](https://arxiv.org/html/2604.21051v1/x3.png)
New research reveals that many software patches offer only superficial changes, leaving residual risks lurking in seemingly secure code.
Science
Beyond Symmetry: A New Look at Quantum Supergroups
Researchers have achieved a complete realization of extended orthosymplectic quantum supergroups, unlocking deeper insights into their intricate mathematical structure.
Agent Pipelines: Uncovering Hidden Security Weaknesses
A new framework systematically exposes developer pitfalls in the security of systems that orchestrate large language model agents.
Securing 5G Location: Defending Against Positioning Attacks
As 5G networks become increasingly vital, ensuring the integrity of location data is paramount, and this review explores methods to protect against sophisticated spoofing and jamming threats.
Mapping the Unknown: Uncertainty in Neutron Star Physics
![The study demonstrates how constraints progressively refine predictions for the mass-radius relation of neutron stars, narrowing the range of possible equations of state-from those based solely on chiral effective field theory, to those incorporating maximum mass limits [latex]M_{max} \leq 2.16^{+0.17}_{-0.15}[/latex], and finally, those informed by astrophysical observations of pulsars like PSR J0030+0451 and gravitational wave events like GW170817, ultimately revealing which equations of state persist under increasing scrutiny.](https://arxiv.org/html/2604.21039v1/plots/CQR-last.png)
A new approach combines statistical rigor with astrophysical modeling to reliably quantify uncertainties in the elusive equation of state governing neutron stars.
Taming Weak Memory: A Formal Approach to Program Verification
Researchers have developed a novel method for rigorously proving the correctness of programs running on modern processors with relaxed memory access rules.
Learning to Trust: Sharpening Aerial Imagery Analysis with Self-Supervised Learning
A new approach to self-supervised learning leverages uncertainty estimation to build more robust representations from aerial imagery, even when data is incomplete or corrupted.
Unmasking AI’s Hidden Risks
New research reveals how interpretability techniques can systematically expose vulnerabilities in even the most advanced large language models.
Root of Trust: Securing IoT and AI Against Hardware Attacks
As connected devices proliferate and artificial intelligence moves to the edge, ensuring hardware integrity is paramount, and this review explores how unique silicon fingerprints can provide a robust defense.
Quantum Simulation Tackles the Strong Force
Researchers are leveraging the Variational Quantum Eigensolver to model the behavior of light nuclei, paving the way for more accurate quantum simulations of nuclear physics.