PreprintNot peer reviewedQuantum ComputingA preprint proposes ancilla-controlled circuits to limit Lie-algebra growth in quantum networksOriginal title: Restricting Trainable Lie-Algebra Growth in Equivariant Quantum Networks via Hierarchical Ancilla-Controlled Subspace ProjectionsResearchers report a quantum circuit design that bounds trainable operator growth while keeping rotation symmetry, with early simulations showing slower gradient decay.Ting Li et al. · Added 1 Nov 2026
PreprintNot peer reviewedQuantum ComputingPreprint proposes new W-state quantum secret sharing protocol tested on real hardwareOriginal title: A Three-Party W-State Quantum Secret Sharing Protocol with X-Gate Encoding and Forbidden-Outcome DetectionA preprint describes a three-party quantum secret sharing scheme using W states and reports error rates varying across quantum processors.Timur R. Teregulov et al. · Added 1 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports multimodal models struggle to read physical fields in new benchmarkOriginal title: PhysFieldBench: Can Multimodal Models Understand Physical Fields?PhysFieldBench tests whether multimodal AI models can infer physics from field images, and finds current models perform poorly.Yuezhou Ma et al. · Added 1 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports task type shapes speculative decoding acceptance more than tree depthOriginal title: Acceptance Dynamics Across Cognitive Domains in Speculative DecodingA preprint finds chat prompts accept more speculative tokens than code or maths, with task type mattering more than tree depth.Seifeldin Abdellatif · Added 1 Nov 2026
PreprintNot peer reviewedQuantum ComputingPreprint describes new quantum error-correcting codes with transversal A7 gate symmetryOriginal title: From Steane to A7: Quantum Codes from Invariant StatesA preprint reports a method for building seven-ququart quantum codes that natively support a non-Clifford gate set.Ian Teixeira · Added 30 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report a membership inference attack on federated prompt tuningOriginal title: Leveraging Soft Prompts for Privacy Attacks in Federated Prompt TuningA preprint describes PromptMIA, an attack where a malicious server infers client data membership via adversarial soft prompts in federated prompt tuning.Quan Minh Nguyen et al. · Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint links hydrogen tunnelling in niobium to lattice motion, not just particle massOriginal title: Transition-state lattice modes and the breakdown of adiabatic tunneling for hydrogen and deuterium in bcc NbA preprint reports that hydrogen and deuterium tunnelling in niobium depends on coupled lattice motion rather than adiabatic separation.P. Graham Pritchard et al. · Added 30 Oct 2026
PreprintNot peer reviewedFrom NVIDIAArtificial IntelligenceResearchers report a cheaper way to fine-tune diffusion models for reward signalsOriginal title: Scaling Reinforcement Learning for Diffusion Models via Velocity MatchingA preprint proposes fine-tuning diffusion and video models by matching velocities weighted by reward, skipping likelihood estimation entirely.Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report faster algorithms for balanced, frustration-free quantum SAT instancesOriginal title: Degree Balance as a Fine-Grained Complexity Boundary for Quantum SATA preprint reports speedups for quantum satisfiability when constraints are evenly spread, with matching hardness limits.Atsuya Hasegawa et al. · Added 30 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceBenchmark tests whether LLM agents can reconstruct full APT attack chains from logsOriginal title: CyberClear: A Benchmark for LLM Agent Systems on APT Attack Chain ProvenanceA preprint introduces CyberClear, a benchmark and harness for testing LLM agents on reconstructing APT attack chains from raw security logs.Qi Chen et al. · Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingA preprint links input angle distributions to barren plateaus in quantum circuitsOriginal title: Unflattening by Flattening — How Input Distributions Shape Output Variance in Angle-Encoded CircuitsThe researchers report how the way input data is encoded as angles shapes output variance and barren plateaus in parameterized quantum circuits.Melvin Strobl et al. · Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers propose an exact phase-space measure of magic for qubit statesOriginal title: A Phase-Space Geometric Measure of Magic in Qubit SystemsA preprint works out an exact geometric measure of magic in qubit systems and its frame dependence.Soumyojyoti Dutta et al. · Added 30 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceStudy finds LLMs conform more to in-group peers, less to out-groupsOriginal title: LLMs Trust Their Own: Identity-Dependent Conformity in Multi-Agent SystemsA preprint reports that language models shift how much they conform to incorrect peers depending on whether those peers share their assigned social identity.Liron Soffer et al. · Added 30 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers test how well LLMs resist confidently stated misinformation in health, finance, lawOriginal title: AdversaRiskQA: An Adversarial Factuality Benchmark for High-Risk DomainsA preprint introduces a benchmark testing whether LLMs detect and correct confidently phrased misinformation in high-risk domains.Adam Szelestey et al. · Added 30 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report LLMs’ own safety judgement can be weaponised to bypass refusalsOriginal title: Safety Paradox: How Enhanced Safety Awareness Leaves LLMs Vulnerable to Posterior AttackA preprint describes a jailbreak that exploits a model's own safety-classification skill, finding better safety judgement means greater vulnerability.Long P. Hoang et al. · Added 30 Oct 2026
PreprintNot peer reviewedFrom QpiAIQuantum ComputingPreprint tests quantum and multireference methods on a Ru single-atom N2 hydrogenation stepOriginal title: Active-Space Quantum Simulation of N$_2$ Hydrogenation at a Ru Single-Atom Site on Ru(0001)A preprint combines periodic DFT with active-space quantum and multireference calculations to study a tricky nitrogen hydrogenation step on a single Ru atom.Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingA preprint reports a quantum method for estimating non-Hermitian pseudospectraOriginal title: Quantum Estimation of Non-Hermitian PseudospectraA preprint proposes a quantum algorithm to test pseudospectrum membership for non-Hermitian operators and tests it on a trapped-ion computer.Gengzhi Yang et al. · Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint proposes quantum memory that builds frequency-bin qubits using radio pulsesOriginal title: Deterministic synthesis and processing of frequency-bin qubits in a macroscopically coherent quantum memoryA preprint outlines a theoretical quantum memory design that could shape photonic qubits via radio-frequency control instead of optical pulse shaping.S.A.Moiseev · Added 30 Oct 2026
PreprintNot peer reviewedArtificial IntelligencePreprint proposes an LLM agent that designs reward functions more efficientlyOriginal title: Beyond Scripted Search: Sample-Efficient Reward Discovery via Agentic Black-box OptimizationA preprint reports an agent-driven search system that finds reinforcement learning reward functions using far fewer costly training runs.Minghao Li et al. · Added 30 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint proposes hardware protocol to measure a channel’s entanglement-breaking indexOriginal title: Measuring the entanglement-breaking index of a qubit collision channel on a quantum processor: protocol and worked exampleA preprint describes a protocol for directly measuring how many rounds a qubit channel needs before it destroys entanglement, tested in simulation.Eran Kopel · Added 30 Oct 2026
