Journal articleReview status unknownArtificial IntelligenceSelf-supervised method trains language models to check medical text for errorsOriginal title: MedVAL: Toward Expert-Level Medical Text Validation with Language ModelsA journal article presents MedVAL, a distillation method that trains language models to validate medical text outputs without physician labels.Asad Aali et al. · Added 7 Nov 2026 · Released Feb 2026
PreprintNot peer reviewedFrom KlingAI ResearchArtificial IntelligenceResearchers report a scale-selection method for NVFP4 quantisation of large language modelsOriginal title: Replay the Curvature: Accurate and Scalable NVFP4 Quantization for Large Language Model InferenceA preprint describes Schur Replay, a quantisation technique said to preserve near-BF16 accuracy while cutting quantisation time sharply.Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers pinpoint a late-layer trigger behind LLM prompt-injection complianceOriginal title: Where Do LLMs Decide to Break the Rules? Mechanistic Localization of Prompt Injection ComplianceA preprint reports that language models decide to obey injected instructions only in a late-layer bottleneck, not where the attack is first encoded.Rui Wen et al. · Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report a method to align reasoning and responses for multilingual LLM safetyOriginal title: actr: aligning thoughts and responses for multilingual safety in reasoning llmsA preprint proposes ACTR, which targets specific neurons to make reasoning LLMs act on their own safety judgements across languages.Xianhui Zhang et al. · Added 6 Nov 2026
PreprintNot peer reviewedFrom XiaomiArtificial IntelligenceNew benchmark tests AI agents across text, images, and multi-step tool useOriginal title: OmniStepBench: Deployment-Aware Agent Evaluation across Modalities and Task ComplexityA preprint introduces OmniStepBench, testing commercial and open-weight models on tasks spanning modalities and execution complexity.Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint reports a method to curb modality-gap drift in continually trained CLIP modelsOriginal title: Stabilizing the Modality Gap: Null-Space Projected Continual Learning for Vision-Language ModelsResearchers propose FPCD, combining null-space gradient projection with contrastive distillation to stabilise vision-language alignment during continual learning.Tiantian Peng et al. · Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint names ‘acquisition collapse’ in costly LLM signal routing decisionsOriginal title: RAISE: Diagnosing Acquisition Collapse in Costly LLM SignalsResearchers propose RAISE, a pre-routing audit for testing whether a costly LLM signal's value is predictable enough to justify selective use.Ying Yuan et al. · Added 6 Nov 2026
PreprintNot peer reviewedFrom Ant GroupArtificial IntelligenceA preprint proposes letting frozen LLMs learn orchestration skills without retrainingOriginal title: Skill-MAS: Evolving Meta-Skill for Automatic Multi-Agent SystemsResearchers report a method that lets frozen large language models improve multi-agent orchestration by refining a written 'Meta-Skill' rather than updating model weights.Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceA preprint maps which quantization choices make LLMs practical to prove with zero-knowledge proofsOriginal title: Bits Under ZK-LLM: Evaluating Zero-Knowledge-Friendly Quantization for Verifiable Private LLM InferenceResearchers report that activation precision and lookup-table design, not just bit-width, drive utility loss and proving cost in zero-knowledge LLM inference.Taeung Yoon et al. · Added 6 Nov 2026
PreprintNot peer reviewedFrom NVIDIAArtificial IntelligenceNVIDIA reports open MoE hybrid model aimed at long agentic tasksOriginal title: Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic ReasoningNVIDIA describes Nemotron 3 Ultra, a large mixture-of-experts hybrid model the authors say boosts inference speed while matching rival accuracy.Added 6 Nov 2026 · Released Jun 2026
PreprintNot peer reviewedArtificial IntelligencePreprint finds multi-turn chat badly degrades retrieval-augmented generation accuracyOriginal title: Lost in Conversation or Lost in Translation? Diagnosing Multi-Turn Degradation in RAGA preprint reports that spreading questions across conversation turns makes RAG systems both less accurate and less reliable.Pranav Handa et al. · Added 6 Nov 2026
PreprintNot peer reviewedFrom Anyscale / SnowflakeArtificial IntelligenceResearchers report a CPU offload method that cuts GPU idle time in LLM fine-tuningOriginal title: ZenFlow: Enabling Stall-Free Offloading for LLM TrainingA preprint describes ZenFlow, which prioritises important gradients to reduce GPU stalls during offloaded LLM training.Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligencePreprint proposes a universal action layer to help AI agents handle messy tool ecosystemsOriginal title: AnyAct: Universal Action for Self-Evolving AgentsA preprint reports a unified action layer, AnyAct, that filters and ranks tools dynamically to help agents cope with huge, changing tool ecosystems.Lingrui Xu et al. · Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers test LLMs’ social intelligence using real Reddit persuasion debatesOriginal title: PerSo: Investigating Facets of LLMs' Social Intelligence in Real-World Persuasion ScenariosA preprint introduces PerSo, a benchmark using real ChangeMyView debates to test whether LLMs can read minds and pick persuasive arguments.Ali Zamani et al. · Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report back-translation boosts LLM translation for low-resource languagesOriginal title: Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMsA preprint reports that back-translating monolingual text improves LLM translation into low-resource languages more than extra pretraining or parallel data.Divyanshu Bathla et al. · Added 6 Nov 2026
PreprintNot peer reviewedFrom Maro Systems / AMDArtificial IntelligenceResearchers propose frozen-backbone method to curb forgetting in continual LLM fine-tuningOriginal title: CRAFT: Forgetting-Aware Intervention-Based Adaptation for Continual LearningA preprint describes CRAFT, which adapts language models to new tasks via representation edits rather than weight changes, aiming to reduce forgetting.Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report a training method to curb context-driven factual errors in LLM reasoningOriginal title: Reasoning Error from Known Fact: Step-Level Self-Consistency Group Relative Policy Optimization for LLMsA preprint presents SSC-GRPO, which uses self-consistency across rollouts to reward correct reasoning steps and reduce hallucinations.Xiaomeng Hu et al. · Added 6 Nov 2026
PreprintNot peer reviewedFrom FactiverseArtificial IntelligenceStudy finds five LLM factuality alignment methods fail to beat base modelOriginal title: Do Factuality Methods Make LLMs More Factual? Reproducibility and Fair ComparisonA preprint reports that none of five tested factuality alignment methods for LLMs reliably beat the untrained base model on a shared benchmark.Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report a mixed-precision method compressing LLM weights to 1.6 bits eachOriginal title: Squeeze10-LLM: Squeezing LLMs' Weights by 10 Times via a Staged Mixed-Precision Quantization MethodA preprint describes Squeeze10-LLM, a staged quantization method claimed to compress LLM weights tenfold with reduced accuracy loss.Qingcheng Zhu et al. · Added 6 Nov 2026
PreprintNot peer reviewedArtificial IntelligenceResearchers report internal model signals can detect training data in fine-tuned LLMsOriginal title: inLoRA-MINT: Internal Representations for Auditing Training Data in Domain-adapted LLMsA preprint reports a method using internal model activations to audit whether specific text was used to fine-tune LLMs with LoRA.Gonzalo Mancera et al. · Added 6 Nov 2026
