Journal articleReview status unknownArtificial IntelligenceA preprint reports no demographic bias detected in an LLM agent action benchmarkOriginal title: AgentFairBench: Do LLM Agents Discriminate When They Act?The authors present a benchmark for demographic disparity in LLM agent actions and report no bias signal for one tested model once noise is corrected.Triveni Morla et al. · Added 4 Oct 2026 · Released Jan 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report LLMs struggle to translate security requests into exact Kali Linux commandsOriginal title: KaliBench: A Fine-Grained Benchmark for Cybersecurity Tool Use on Kali Linux with Runtime-Free Verifiable RewardsA preprint introduces KaliBench, a benchmark showing open-weight models often fail to turn plain-English security requests into correct CLI commands.Pengfei Li et al. · Added 4 Oct 2026
PreprintNot peer reviewedFrom Google Quantum AIQuantum ComputingResearchers report proof of exponential quantum edge in classical data tasksOriginal title: Exponential quantum advantage in processing massive classical dataA preprint claims small quantum machines can beat exponentially larger classical ones at classification and dimension reduction on streamed data.Added 4 Oct 2026
PreprintNot peer reviewedFrom FuriosaAIArtificial IntelligenceResearchers report faster RL rollouts via self-speculative decodingOriginal title: EfficientRollout: System-Aware Self-Speculative Decoding for RL RolloutsA preprint describes EfficientRollout, a system-aware self-speculative decoding method that speeds up LLM reinforcement-learning rollouts without hurting final quality.Added 4 Oct 2026
Journal articleReview status unknownQuantum ComputingStudy finds common continuous parity-measurement schemes corrupt quantum error correctionOriginal title: Obstacles to Continuous Quantum Error Correction via Parity MeasurementsThe authors report that standard continuous parity-measurement protocols leak logical information, favouring erasure-based encodings instead.Anton Halaski et al. · Added 4 Oct 2026 · Released Mar 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers build a fine-grained Chinese LLM safety guardrail trained on new datasetsOriginal title: CHILLGuard: Towards Fine-Grained Chinese LLM Safety Guardrail with Scalable Data Construction and Model-aware Preference AlignmentA preprint introduces CHILLGuard, a Chinese content-safety guardrail with a 31-category taxonomy and preference-tuned training.Wenbo Yu et al. · Added 4 Oct 2026
PreprintNot peer reviewedQuantum ComputingResearchers report new dataset to test LLMs on hardware-level quantum code generationOriginal title: QASM-Eval: A Dataset to Train and Evaluate LLMs on OpenQASM-3 Beyond Quantum CircuitsA preprint introduces QASM-Eval, a benchmark and training set for LLMs writing hardware-facing OpenQASM 3 code, with fine-tuning showing large gains.Zhenxiao Fu et al. · Added 4 Oct 2026
Journal article✓ Peer reviewedArtificial IntelligenceBenchmark finds LLM agents widely vulnerable to hidden tool-metadata attacksOriginal title: MCPTox: A Benchmark for Tool Poisoning Attack on Real-World MCP ServersA new benchmark shows many LLM agents can be tricked by malicious instructions hidden in tool descriptions, not just tool outputs.Zhiqiang Wang et al. · Added 4 Oct 2026 · Released Mar 2026
Journal articleReview status unknownQuantum ComputingResearchers propose a parallelisable quantum algorithm for quantum chemistry on noisy hardwareOriginal title: Parallel Quantum Chemistry on Noisy Intermediate-Scale Quantum ComputersA method splits molecular energy calculations into smaller linked pieces to cut qubit needs and run on today's noisy quantum computers.Robert Schade et al. · Added 4 Oct 2026 · Released Aug 2022
PreprintNot peer reviewedArtificial IntelligenceResearchers propose an energy-landscape theory for why jailbreaks work on diffusion language modelsOriginal title: Why Jailbreaks Succeed in Diffusion Language Models: An Energy Landscape AnalysisA preprint frames dLLM safety as an energy barrier and derives three training-free signals that detect known jailbreak attacks.Thong Bach et al. · Added 4 Oct 2026
Journal articleReview status unknownFrom Equal1 LaboratoriesQuantum ComputingResearchers report silicon spin qubits working at higher temperatures than beforeOriginal title: Persistence of Entangled States and High Fidelity Quantum Gate Operations in Si/SiGe Spin Qubits at High TemperatureA study of SiGe spin qubits reports high-fidelity single- and two-qubit operations up to 750 mK, the warmest yet reported for this platform.Added 4 Oct 2026 · Released Mar 2026
Journal articleReview status unknownArtificial IntelligenceResearchers report an AI agent framework for controlling optical network hardware via natural languageOriginal title: Agentic AI for Scalable and Robust Optical Systems ControlA framework called AgentOptics lets large language models control real optical devices through a standard tool protocol, with authors reporting high task success rates.Zehao Wang et al. · Added 4 Oct 2026 · Released Jan 2026
PreprintNot peer reviewedQuantum ComputingResearchers report quantum simulation of a lattice gauge theory using trapped-ion qubits and oscillatorsOriginal title: Real-Time Observation of Aharonov-Bohm Interference in a $\mathbb{Z}_2$ Lattice Gauge Theory on a Hybrid Qubit-Oscillator Quantum ComputerA preprint reports a hybrid qubit-oscillator quantum device simulating gauge field dynamics and observing Aharonov-Bohm interference.S. Saner et al. · Added 4 Oct 2026 · Released Jul 2025
Journal article✓ Peer reviewedQuantum ComputingStudy maps readout sweet spots for spin-orbit qubits using charge sensingOriginal title: Readout sweet spots for spin qubits with strong spin-orbit interactionResearchers model how charge-sensor back-action degrades spin qubit readout and locate configurations closest to ideal measurement.Domonkos Svastits et al. · Added 4 Oct 2026 · Released Mar 2026
Journal articleReview status unknownArtificial IntelligenceResearchers report widespread bias in LLM-based judges and propose a debiasing training methodOriginal title: Toward Robust LLM-Based Judges: Taxonomic Bias Evaluation and Debiasing OptimizationA new benchmark finds LLM judges remain biased across 12 types, and a training method reduces this while keeping accuracy.Hongli Zhou et al. · Added 4 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report security flaws in world models used by AI agentsOriginal title: False Prophets: On the Security of World Models in Agentic SystemsA preprint finds text-based world models used by AI agents can be tricked into dangerous mispredictions up to 95% of the time.Erik Imgrund et al. · Added 4 Oct 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report encode-aware serving design boosts multimodal LLM throughputOriginal title: EAServe: Encode-Aware Disaggregated Serving for Multimodal Large Language ModelsA preprint proposes treating the Encode stage as the control point for serving multimodal LLMs, reporting large goodput gains over existing systems.Kunxiong Zhu et al. · Added 4 Oct 2026
Journal articleReview status unknownArtificial IntelligenceStudy tests ways to guide LLMs toward writing and fixing more secure codeOriginal title: Guiding AI to Fix Its Own Flaws: An Empirical Study on LLM-Driven Secure Code GenerationResearchers examine whether self-written vulnerability hints and detailed feedback help large language models avoid or repair insecure code.Hao Yan et al. · Added 4 Oct 2026
Journal articleReview status unknownQuantum ComputingResearchers report decoding gains for quantum codes using limited error-rate dataOriginal title: Enhancing Decoding Performance using Efficient Error LearningA study reports that feeding a decoder a small set of measured error rates sharply cuts logical error rates in concatenated quantum codes.Pavithran Iyer et al. · Added 4 Oct 2026 · Released Jul 2025
Journal article✓ Peer reviewedFrom Red Blue QuantumQuantum ComputingResearchers report faster protocol for preparing bosonic quantum error-correction statesOriginal title: Fast Bosonic Control via Multiphoton Qubit-Oscillator InteractionsA peer-reviewed study proposes multiphoton qubit-oscillator interactions that the authors say cut preparation times for bosonic codewords.Added 4 Oct 2026
