PreprintNot peer reviewedArtificial IntelligenceResearchers propose Bayesian method to cut cost of repeated LLM evaluation runsOriginal title: Sequential Bayesian Evaluation of Large Language Model BehaviorA preprint presents a Bayesian sequential sampling method that reduces the number of generations needed to evaluate stochastic LLM behaviour.Saatvik Kher et al. · Added 28 Oct 2026
Journal articleReview status unknownQuantum ComputingStudy tests quantum machine learning for classifying neutrino telescope eventsOriginal title: Neutrino Telescope Event Classification on Quantum ComputersThe authors report that quantum classifiers can distinguish neutrino event types with accuracy similar to classical methods.Pablo Rodriguez-Grasa et al. · Added 28 Oct 2026 · Released Mar 2026
PreprintNot peer reviewedFrom Ant GroupArtificial IntelligenceResearchers report clinical AI agents falter when patient records lack supporting evidenceOriginal title: EHR-RobustGym: Benchmarking and Training Agents for Robust Clinical ReasoningA preprint introduces a benchmark showing clinical LLM agents struggle to detect when hospital records lack evidence to answer a query.Added 28 Oct 2026
PreprintNot peer reviewedFrom FuriosaAIArtificial IntelligenceResearchers simulate flash-backed memory to speed up LLM servingOriginal title: Characterizing High Bandwidth Flash for LLM ServingA preprint reports that adding high-bandwidth flash with smarter scheduling can cut LLM serving time and energy use in simulation.Added 28 Oct 2026
Journal articleReview status unknownQuantum ComputingStudy tests a quantum algorithm for discrete portfolio optimisation under noiseOriginal title: Quantifying the advantages of applying quantum approximate algorithms to portfolio optimisationThe authors design an end-to-end QAOA pipeline for discrete portfolio selection and find noise, not algorithm design, sets its limits.Haomu Yuan et al. · Added 28 Oct 2026 · Released Mar 2026
PreprintNot peer reviewedQuantum ComputingA preprint reports a sharp timing jump in quantum simulation on Apple M4 Pro memoryOriginal title: A Controlled Study of Memory Hierarchy Transitions in Quantum Circuit Simulation on Apple M4 Pro Unified Memory ArchitectureA preprint finds a reproducible timing cliff at 28 to 29 qubits in quantum circuit simulation on Apple's unified memory chips.Gyan Pratipat · Added 28 Oct 2026
PreprintNot peer reviewedFrom Amazon, MicrosoftArtificial IntelligenceA preprint reports training on aligned data can still cause narrow misalignment elsewhereOriginal title: Aligned Data Can Induce Misalignment via Context ConfusionResearchers describe context confusion, where fine-tuning on data that is aligned in one domain makes models misaligned in a different, related domain.Added 28 Oct 2026
PreprintNot peer reviewedQuantum ComputingA preprint reports a near-optimal quantum algorithm for estimating Chern numbersOriginal title: Chern-number estimation: nearly optimal quantum algorithm and provable quantum speedupResearchers describe a quantum algorithm for topological invariant estimation and prove matching limits on its speed, plus computational hardness results.Soichiro Imamura et al. · Added 28 Oct 2026
Journal articleReview status unknownFrom Chunghwa TelecomQuantum ComputingResearchers propose hash-based anonymous certificate scheme for post-quantum securityOriginal title: Post-Quantum Cryptography Anonymous Scheme — PQCWC: Post-Quantum Cryptography Winternitz-ChenA study proposes PQCWC, a hash-based anonymous certificate scheme using Winternitz signatures and a new butterfly key expansion method for quantum safety.Added 28 Oct 2026 · Released Sep 2024
PreprintNot peer reviewedQuantum ComputingResearchers report single-server quantum private database retrieval using only NISQ devicesOriginal title: One-Way Quantum Symmetric Private Information Retrieval Protocol From A Single Database Server Using NISQ DevicesA preprint describes a one-way quantum protocol letting a client fetch one database bit privately, without needing multiple servers or quantum memory.Xiang Zou et al. · Added 28 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report compact LLM self-trains for graph reasoning without external teachersOriginal title: GraphCert: Bootstrap Agentic Graph Reasoning with Certified Evidence RubricsA preprint describes GraphCert, a framework letting a small language model generate and verify its own training data for reasoning over knowledge graphs.Weiqi Jiang et al. · Added 28 Oct 2026
PreprintNot peer reviewedFrom MemoraX AIArtificial IntelligenceA preprint proposes a benchmark to separate real skill reuse from code copying in AI agentsOriginal title: EngramBench: A Capability-Grounded Benchmark for Skill-Evolution HarnessesA preprint introduces EngramBench to test whether coding agents genuinely transfer skills rather than reuse memorised code.Added 28 Oct 2026
PreprintNot peer reviewedArtificial IntelligencePreprint compares six RAG pipelines for scientific QA, finds late interaction winsOriginal title: Re-ranking and Late Interaction Drive Retrieval Quality: A Controlled Comparison of RAG Strategies for Scientific Question AnsweringA preprint reports that ColBERTv2's late-interaction retrieval beat five single-vector RAG strategies on scientific question answering.Bhagyesh Rathi et al. · Added 28 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report a magic Mpemba effect in localized quantum dynamics on IBM hardwareOriginal title: Observing the magic Mpemba effect in localized dynamics on a digital quantum computerA preprint describes states with less initial quantum magic overtaking more magical ones, observed on superconducting quantum processors.Han-Ze Li et al. · Added 28 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports serverless scheme to curb MoE LLM inference stragglersOriginal title: MoEless: Efficient MoE LLM Serving with Serverless ExpertsResearchers report a serverless framework that predicts and rebalances overloaded MoE experts to cut inference latency and cost.Hanfei Yu et al. · Added 26 Oct 2026
PreprintNot peer reviewedFrom China Electronic Technology Cyber SecurityQuantum ComputingA preprint reports a feedback quantum optimizer with fewer measurements, similar depthOriginal title: Feedback-based quantum optimization with low depth and measurementResearchers propose BLS-FALQON, which they report cuts measurement count versus a prior low-depth FALQON variant while keeping similar circuit depth.Added 26 Oct 2026
PreprintNot peer reviewedFrom MetaArtificial IntelligencePreprint finds post-training often narrows agentic task coverage, not just boosts accuracyOriginal title: Sharpening Tax in Post-TrainingA preprint argues RL post-training sharpens LLM behaviour towards reliability while shrinking the range of agentic tasks it can solve.Added 26 Oct 2026
PreprintNot peer reviewedQuantum ComputingA preprint reports a faster way to simulate hybrid quantum systems without Fock truncationOriginal title: Efficient Simulation of Hybrid Continuous- and Discrete-Variable Quantum Circuits via Gaussian DecompositionsResearchers describe a simulation method for combined discrete- and continuous-variable quantum circuits that avoids costly Fock-space truncation.Kazufumi Tanji et al. · Added 26 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report block-wise quantum PDE solver beats global ansatz on rough regionsOriginal title: Block-Wise Variational Quantum Algorithms for PDEs with Interface Penalty ConstraintsA preprint proposes splitting PDE domains into quantum sub-blocks with interface penalties, reporting gains over global ansatz methods in some regimes.Hangran Jie et al. · Added 26 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report faster quantum method for equilibrium correlation functionsOriginal title: Efficient Calculation of Equilibrium Correlation FunctionsA preprint describes a clock-register quantum algorithm that computes correlation functions for all time points with one shared circuit.Yizhi Shen et al. · Added 26 Oct 2026
