Can AI Reason About Code Concurrency?
A new approach combines the power of large language models with formal verification to tackle the notoriously difficult problem of finding bugs in concurrent programs.
Science
Unlocking QCD’s Secrets: A New Calculation for Particle Interactions
![The study dissects the intricacies of non-singlet splitting functions - [latex]P_{ns}^{(3)+}, P_{ns}^{(3)-}, P_{ns}^{(3)s}[/latex] - at [latex]n_f = 4[/latex], revealing how approximations, when contrasted with the full calculation, illuminate the inherent complexities within quantum chromodynamics and the subtle interplay between theoretical models and observed phenomena.](https://arxiv.org/html/2604.09534v1/x19.png)
Researchers have completed a complex calculation refining our understanding of how quarks and gluons interact within protons and neutrons.
Siloed Searches: Transforming Queries for Cross-Organization Data Access
![The system architecture preserves data sovereignty through a five-phase workflow across organization-specific vector spaces, and achieves computational isolation of queries via a multi-stage transformation-[latex]vector2Trans[/latex]-employing permutation, cryptographic blinding, a bounded non-linearity [latex]f_{\beta}[/latex], orthogonal rotation [latex]W[/latex], and L2 normalization while retaining retrieval utility.](https://arxiv.org/html/2604.09541v1/Graph/overview.png)
A new approach to Retrieval-Augmented Generation shifts the focus from protecting documents to transforming queries, enabling secure and efficient data sharing between organizations.
Safeguarding Robots with Uncertainty: A Probabilistic Approach
This research introduces a new framework for ensuring the safe operation of robotic systems even when facing imperfect state information and external disturbances.
Speeding Up AI Answers: A Smarter Way to Cache Information
New research introduces a system that dramatically improves the efficiency of retrieving and processing information for large language models, leading to faster and more cost-effective AI responses.
Marking the Machine: Securely Embedding Data in AI Text
New research establishes the theoretical limits of how much information can be reliably hidden within the outputs of large language models.
Bouncing Back to Isotropy: How Quantum Effects Tame the Early Universe
![The study demonstrates that, within a Bianchi I universe populated by radiation, the Hubble horizon-defined as [latex]L_{H} \equiv (aH)^{-1}[/latex]-exhibits qualitatively consistent behavior across varying initial conditions established at [latex]t=0[/latex], as evidenced by the consistent evolution of [latex] \pi_{i}(0)[/latex] and [latex] c_{i}(0)[/latex]-a finding expressed in Planck units.](https://arxiv.org/html/2604.08912v1/x4.png)
New research suggests that quantum effects within a modified loop quantum cosmology framework can effectively suppress early universe anisotropies, paving the way for a remarkably uniform cosmos.
Beyond the Semicircle: New Limits for Polynomial Intersection
A new study reveals unexpectedly efficient solutions to the Optimal Polynomial Intersection problem, challenging previously established performance bounds for quantum algorithms.
Untangling Quantum Errors with Majorana’s Ghost
![The syndrome distribution of the honeycomb-lattice toric code, perturbed by [latex]ZZ[/latex]-type coherent errors, is recast as a non-interacting, monitored circuit acting on a Majorana chain with distinct [latex]AA[/latex] and [latex]BB[/latex] sublattices-a transformation belonging to symmetry class D-where the statistical model’s disordered structure maps directly to circuit blocks evolving between discrete timesteps.](https://arxiv.org/html/2604.08650v1/x17.png)
New research reveals a surprising connection between decoding topological quantum codes and the behavior of monitored Majorana fermions, offering insights into the fundamental limits of quantum error correction.
Decoding Security: New Insights into the CROSS Signature Scheme
A new analysis examines the mathematical underpinnings of the CROSS digital signature, assessing its resilience against attacks by reducing its core problem to well-studied cryptographic challenges.