Research & Innovation

Global: Spatial-Agent Enables Executable Geospatial Reasoning in Large Language ModelsA team of researchers introduced Spatial-Agent, an artificial‑intelligence system designed to answer geospatial questions by converting natural‑language queries into executable workflows.…

Global: New arXiv Review Maps AI’s Expanding Role in Creative IndustriesScope of the ReviewThe authors of a recent arXiv preprint (arXiv:2501.02725) released in January 2025 present a systematic review of…

Global: AgentDrive Benchmark Released to Evaluate Autonomous Driving Language ModelsResearchers have unveiled AgentDrive, a large‑scale benchmark comprising 300,000 driving scenarios generated by large language models (LLMs), intended to support the…

Global: GRIP Framework Improves Machine Unlearning for Mixture-of-Experts Language ModelsIn January 2026, the authors of a newly posted arXiv preprint introduced Geometric Routing Invariance Preservation (GRIP), an algorithm‑agnostic adapter designed…

Global: New Metrics Reveal Intention Collapse Dynamics in Large Language ModelsResearchers have presented three model‑agnostic metrics designed to quantify the relationship between a language model’s internal intention space and its…

Global: Multi-Agent Graph-Enhanced Knowledge Tracing Advances Student Performance PredictionIn January 2026, a team of researchers released a new preprint on arXiv introducing Multi-Agent Graph-Enhanced Knowledge Tracing (MAGE-KT), a framework designed…

Global: AutoSciDACT Introduces Robust Novelty Detection Pipeline for Scientific DatasetsIn October 2025, a team of researchers announced AutoSciDACT, a new pipeline designed to improve novelty detection in large scientific datasets…

Global: New NSED Protocol Enables Small Model Ensembles to Rival Large-Scale AI SystemsResearchers announced the N‑Way Self‑Evaluating Deliberation (NSED) protocol in a January 2026 arXiv preprint, describing a runtime mixture‑of‑models…

Global: Reasoning-Enhanced Language Models Show Greater Robustness on Theory of Mind TasksThree researchers—Ian B. de Haan, Peter van der Putten, and Max van Duijn—submitted a study to arXiv on January…

Global: New Algorithm Achieves Optimal I/O Complexity for Small-Cache AttentionResearchers from several institutions have presented a new algorithm that reaches theoretical I/O complexity limits for attention mechanisms when operating with…