PreprintNot peer reviewedQuantum ComputingResearchers map when compiling quantum circuits from Clifford templates is easy or hardOriginal title: Computational Complexity of Clifford Template Compilation: Are Quantum Computers Useful for Compiling Quantum Circuits?A preprint charts sharp complexity shifts in choosing repetition counts for Clifford templates, from classical ease to NP-completeness and quantum tractability.Keisuke Fujii · Added 29 Oct 2026
PreprintNot peer reviewedFrom MicrosoftArtificial IntelligenceResearchers report an activation-steering defence against indirect prompt injection in LLM agentsOriginal title: CounterSteer: Suppressing Indirect Prompt Injection with Activation SteeringA preprint describes CounterSteer, a per-model steering vector that cuts prompt-injection attacks with little loss of task quality.Added 29 Oct 2026
PreprintNot peer reviewedArtificial IntelligencePreprint reports fine-tuning on search-recoverable failures boosts single-sample LLM accuracyOriginal title: Learning from the Gap Between Pass@K and Pass@1A preprint proposes training only on cases a model misses once but solves via search, lifting single-shot accuracy.Xuan Liu et al. · Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint reports quantum kernel tuned to variable interactions aids fraud detectionOriginal title: When Similarity Is Interaction-Driven: Quantum Kernels for Regime-Sensitive LearningResearchers report a quantum kernel built around interacting variables outperforms standard kernels on synthetic and fraud-detection data.Hanqiu Peng et al. · Added 29 Oct 2026
PreprintNot peer reviewedFrom SalesforceArtificial IntelligenceResearchers report single-token LLM method for fast, accurate classificationOriginal title: Koa-action: Fast and Consistent Structured Decision Making with Generative LLMsA preprint describes Koa-action, which turns classification into a one-step token prediction to cut latency while keeping accuracy competitive with frontier LLMs.Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingA preprint reports a non-variational quantum walk algorithm for optimisation problemsOriginal title: Sampled-Based Guided Quantum Walk: Non-variational quantum algorithm for combinatorial optimizationResearchers propose SamBa-GQW, a quantum walk guided by classical sampling that skips the variational optimizer step used by QAOA.Ugo Nzongani et al. · Added 29 Oct 2026
PreprintNot peer reviewedFrom GraphcoreArtificial IntelligenceA preprint reports a training-free recipe for retrieval from knowledge graphs at Wikidata scaleOriginal title: UltRAG: a Universal Simple Scalable Recipe for Knowledge Graph RAGResearchers describe UltRAG, a training-free system combining LLM query generation with neural graph query execution, reportedly matching Wikidata-scale performance.Added 29 Oct 2026
PreprintNot peer reviewedFrom Beavernets TechnologiesQuantum ComputingResearchers report a memory-bounded Pauli decomposition method for qubits and quditsOriginal title: Pauli Decomposition by Character Theory: A Memory-Bounded Algorithm for Qubits and QuditsA preprint describes paulikit, a Pauli decomposition tool the authors say keeps memory use decoupled from operator size, even for very large sparse systems.Added 29 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report an LLM system that turns expert failure review into reusable policiesOriginal title: From Answers to Policies: Efficient In-Context Learning System through Emulating Expert InvestigationA preprint describes PIHF-MCP, which emulates expert failure investigation to build reusable LLM task policies more cheaply than prior methods.Minh-Ha Nguyen et al. · Added 29 Oct 2026
PreprintNot peer reviewedFrom SAP, Microsoft ResearchArtificial IntelligenceResearchers report an LLM architecture that selects actions directly, not via textOriginal title: Dyad: Extending Large Language Models with Native Typed Decision-MakingA preprint describes Dyad, which adds a typed action encoder to language models so they can pick actions directly rather than generating text for them.Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearcher revisits a decades-old conjecture on entangled-qubit noise and how to test itOriginal title: The Fully Depolarizing Noise Conjecture for Entangled Physical States: A Twenty-Year PerspectiveA preprint revisits a long-standing conjecture that entangled qubits suffer a correlated depolarising noise component, and proposes ways to test it experimentally.Gil Kalai · Added 29 Oct 2026
PreprintNot peer reviewedFrom Samsung Research AmericaArtificial IntelligenceResearchers report a stealthy backdoor attack on RAG systems via poisoned passagesOriginal title: BadRAG: Identifying Vulnerabilities in Retrieval Augmented Generation of Large Language ModelsA preprint describes BadRAG, which hides trigger-activated malicious passages in a RAG knowledge base to steer LLM outputs.Added 29 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports method that fixes dense reward signal issues in LLM reasoning trainingOriginal title: SIPO: Unifying Reinforcement Learning with On-Policy Self-DistillationResearchers propose SIPO, a contrastive self-teacher RL method that reportedly outperforms standard RL and self-distillation on reasoning and coding benchmarks.Zhenrui Yue et al. · Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint reports faster quantum algorithm for decomposing Abelian black-box groupsOriginal title: Improved Quantum Algorithms for Black-Box Abelian Group DecompositionA preprint reports a quantum algorithm that cuts gate counts and qubit use for decomposing finite Abelian black-box groups.Junrong Luo et al. · Added 29 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report a black-box watermark for detecting LLM training on proprietary datasetsOriginal title: Dataset Watermarking with Provable Black-Box DetectionA preprint describes a dataset watermarking method that spots fine-tuning on protected data using only generated text, even when heavily diluted.Pengrun Huang et al. · Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report a full characterisation of optimal fusion strategies for photonic quantum codesOriginal title: Optimal Fusion Strategies for Quantum ComputationA preprint characterises when fusion-based quantum computation can tolerate maximal physical fusion failures using single-qubit codes.Kenneth Goodenough et al. · Added 29 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report LLMs can fine-tune on unreadable synthetic embeddingsOriginal title: Is Human-Readable Text Necessary for Effective LLM Fine-Tuning?A preprint reports that continuous synthetic embeddings, not human-readable text, can match or beat real text for fine-tuning large language models.Jinhao Zhang et al. · Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingPreprint reports new clock-spectroscopy pulse scheme tested on superconducting quantum processorsOriginal title: Fault-tolerant hyper-Ramsey spectroscopy of optical clock transitions with dynamical decouplingA preprint describes dynamically decoupled hyper-Ramsey pulse sequences tested on two superconducting quantum processors to suppress clock-shift errors.B. Ilikj et al. · Added 29 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint describes training that makes models admit when they suspect an evaluationOriginal title: Training LLMs to Verbalize Evaluation AwarenessResearchers report a training method that boosts language models' tendency to say aloud that they think they are being tested, without changing their actual behaviour.Usman Anwar et al. · Added 29 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report cloud-native emulator for hybrid QKD-PQC network testingOriginal title: Hybrid QKD-PQC Network Emulation through Automated and Scalable Cloud-Native OrchestrationA preprint describes Quditto, an emulation platform letting researchers test hybrid quantum-safe networks without physical quantum hardware.Iván Melijosa et al. · Added 29 Oct 2026
