PreprintNot peer reviewedFrom MetaArtificial IntelligenceResearchers propose a trainable harness that lets frozen LLM agents reuse experienceOriginal title: ExpHarness: Model-Agnostic Experience Learning through a Trainable HarnessA preprint describes ExpHarness, an external, trainable experience system that boosts frozen LLM agents without touching their weights.Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report multi-agent system boosts faithfulness of model explanationsOriginal title: MEA: A Reward-Driven Multi-Agent System for Faithful Model ExplanationsA preprint describes MEA, a multi-agent system trained to generate explanations that are verifiably grounded in model behaviour.Yuyang Cheng et al. · Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceDistilling a safety-focused teacher model can make student LLMs easier to jailbreakOriginal title: Response-Based Knowledge Distillation for Multilingual Jailbreak Prevention Unwittingly Compromises SafetyA preprint reports that response-based knowledge distillation meant to boost multilingual safety instead raised jailbreak success rates in several open-source models.Max Zhang et al. · Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports an LLM architecture that reads graph structure nativelyOriginal title: Teaching LLMs to See Graphs: Unifying Text and Structural ReasoningA preprint describes GTLM, which adds graph-aware attention to language models instead of compressing nodes into single tokens.Dario Vajda · Added 25 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint proposes faster pseudorandom Clifford operators using a cryptographic trapdoorOriginal title: Trapdoored Clifford Operators and ApplicationsA preprint reports a way to generate Clifford operators that look random but run much faster if you hold a secret trapdoor.Minki Hhan et al. · Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceStudy finds LLM agents’ numerical signals become structured but behaviour stays unpredictableOriginal title: When Numbers Start Talking: Numerical Signalling and Strategic Behaviour Among LLMsResearchers report that AI agents exchanging numbers develop consistent symbol patterns, yet this does not reliably predict cooperation outcomes across models.Alessio Buscemi et al. · Added 25 Oct 2026
PreprintNot peer reviewedFrom Qualcomm, NVIDIA, Microsoft, Google DeepMindArtificial IntelligenceMLCommons benchmark measures how easily jailbreaks bypass eight open-weight AI modelsOriginal title: MLCommons Jailbreak Benchmark v1.0A preprint from MLCommons reports a new pipeline for testing single-turn jailbreak attacks and measures how much safety degrades across eight open-weight models.Added 25 Oct 2026
PreprintNot peer reviewedFrom CiscoQuantum ComputingResearchers report a new compiler cuts teleportation overhead in distributed quantum computersOriginal title: IRIS: A Compiler for Distributed Quantum SystemsA preprint describes IRIS, a compiler that cuts teleportation counts and latency in distributed quantum systems compared with a prior state-of-the-art compiler.Added 25 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint details method to build quantum codes with targeted transversal Z-rotation gatesOriginal title: Realizing Logical Diagonal Gates via Transversal Physical $Z$-Rotations in CSS CodesA preprint presents a way to extend CSS quantum codes so they transversally support chosen logical diagonal gates.K. Sai Mineesh Reddy et al. · Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceNew benchmark pits humans against LLMs on graph neural network coding tasksOriginal title: GNN-CB: A Graph Neural Network Competition Benchmark for Human and LLM EvaluationA preprint introduces GNN-CB, a competition-style benchmark showing LLMs still trail top human coders on most GNN tasks.Murad Hossen et al. · Added 25 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report framework for simulating quantum dynamics on topological hardwareOriginal title: Scalable Simulation of Quantum Dynamics on Topological Quantum HardwareA preprint outlines a method using anyon braiding to approximate quantum dynamics propagators for molecules and condensed-phase systems.Kritanjan Polley et al. · Added 25 Oct 2026
PreprintNot peer reviewedFrom NVIDIAArtificial IntelligencePreprint reports evaluator model lets frozen robot policies pick better actionsOriginal title: Look Before You Leap: Distilling Tree Search into Action Evaluation for Frozen VLA ModelsA preprint describes SVA, a method that adds a learned action evaluator to frozen VLA robot policies, reportedly boosting real-robot success rates.Added 25 Oct 2026
PreprintNot peer reviewedFrom Alice & BobQuantum ComputingPreprint reports cheaper quantum circuits by splitting where terms act from what they doOriginal title: Efficient Block Encoding of Structured Hamiltonians by Separating Where and WhatA preprint describes block-encoding circuits that separate operator placement from operator action, cutting non-Clifford gate counts for structured Hamiltonians.Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report a context-learning framework doubles AI alignment with human moral judgmentsOriginal title: Morality is Contextual: Learning Interpretable Moral Contexts from Human Data with Probabilistic Clustering and Large Language ModelsA preprint presents COMETH, a modular system that learns context-specific moral judgments and roughly doubles alignment with human ratings versus direct LLM prompting.Geoffroy Morlat et al. · Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports an anonymizer resisting agentic web-search re-identificationOriginal title: LLM Anonymization Against Agentic Re-IdentificationA preprint introduces AURA, a mask-reconstruct method the authors say cuts agentic re-identification while keeping more contextual detail than prior tools.Ziwen Li et al. · Added 25 Oct 2026
PreprintNot peer reviewedQuantum ComputingA preprint reports a new compiler for simulating quantum chemistry on chiplet hardwareOriginal title: QBX: A Compiler for 2-local Qubit Hamiltonian Simulation on Quantum ChipletsResearchers describe QBX, a compiler that tailors Hamiltonian simulation circuits to multi-chiplet quantum hardware, reportedly cutting circuit depth and gate counts.Zikun Li et al. · Added 25 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports LLM agents can re-identify anonymised interviews via web searchOriginal title: AgentDoxx: Agentic Re-identification of Anonymized Text with Web SearchA preprint finds that search-enabled LLM agents can often re-identify anonymised interview transcripts, even after entity masking.Jianing Wen et al. · Added 25 Oct 2026
PreprintNot peer reviewedFrom Impossible, Inc.Artificial IntelligenceResearchers report hybrid attention cuts video generation latency with similar qualityOriginal title: Video DeltaNet: A Video-Native Hybrid Attention for Livestream Video GenerationA preprint describes Video DeltaNet, a hybrid attention method that the authors say speeds up long video generation while keeping quality close to dense attention.Added 25 Oct 2026
PreprintNot peer reviewedFrom OptQCQuantum ComputingPreprint describes PRISM codes enabling full Clifford gates with low-weight stabilizersOriginal title: Constant-rate quantum codes with low-weight stabilizers and full logical Clifford actions via transversal and fold-transversal gatesA preprint presents a quantum code family combining constant rate and full logical Clifford gates via transversal operations alone.Added 25 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint reports fault tolerance scheme resisting near-linear adversarial qubit corruptionOriginal title: Fault-Tolerant Quantum Computation with Adversarial ErrorsA preprint presents a quantum fault-tolerance scheme that the authors say withstands worst-case, correlated corruption of an almost-linear fraction of qudits per step.Nikolas P. Breuckmann et al. · Added 25 Oct 2026
