Research Correspondent
Elena Voss covers model development and the ideas behind new research. Her beat follows the distance between a promising paper and a result that holds up outside the lab. She favors clear explanations, original sources and questions that a benchmark score alone cannot answer.
2,898 published articles · Showing 461–480
A groundbreaking paper from arXiv reveals the emergence of multi-modal AI agents operating as a “full multidisciplinary discovery team” within a semi-autonomous operating system, Rhizome OS-1, specifically designed for small molecule drug discovery [arXiv:2604. 07512]....
Read article →A groundbreaking series of studies, culminating in a paper published today, establishes a profound mathematical isomorphism between ant colony decision-making, various ensemble learning techniques, and the fundamental learning algorithm of deep neural networks: stochastic gradien...
Read article →The frontier of artificial intelligence research is buzzing with dual momentum: a relentless drive to make large language models (LLMs) more efficient and deployable, and a profound push to integrate AI more deeply into the heart of scientific discovery. Recent papers released on...
Read article →A flurry of new research, published today on arXiv, signals a pivotal moment in deep learning, with researchers unveiling innovative approaches to make large AI models more efficient, robust, and capable of nuanced reasoning across diverse applications. These breakthroughs tackle...
Read article →A new research paper introduces LETGAMES, an innovative LLM-powered system designed to automatically generate personalized, interactive narrative games for cognitive training in patients with cognitive impairments. This approach addresses the significant resource intensity tradit...
Read article →Today marks a remarkable moment for AI research as arXiv published a diverse array of cutting-edge papers on April 14, 2026, signaling a profound acceleration in the development of artificial intelligence across numerous disciplines. From foundational algorithmic enhancements to ...
Read article →An artificial intelligence model has achieved a perfect score on an officially disclosed Law School Admission Test (LSAT), marking a significant milestone in AI's capacity for complex logical and analytical reasoning arXiv CS. AI....
Read article →A remarkable cluster of new research papers on arXiv CS. AI, all released on April 14, 2026, signals a pivotal moment in the development of AI agents....
Read article →A significant shift in artificial intelligence is underway, as new research and applications demonstrate AI agents moving beyond simple task execution to deeply understanding and proactively optimizing complex human intent. This evolution is being explored in diverse fields, from...
Read article →A crucial limitation in the development of truly autonomous embodied agents may soon be overcome, thanks to a novel framework named Dejavu. This new approach, detailed in a recent arXiv paper, enables intelligent systems to continuously acquire new knowledge and improve task perf...
Read article →A significant stride has been made in bringing powerful Deep Neural Networks (DNNs) to edge devices with the introduction of MATCHA, a unified deployment framework designed to fully exploit the capabilities of multi-accelerator heterogeneous System-on-Chips (SoCs). This innovatio...
Read article →A new research paper, 'Squeeze Evolve: Unified Multi-Model Orchestration for Verifier-Free Evolution,' reveals a promising approach to overcome two significant challenges in AI development: the computational waste of high-cost models and the tendency for evolutionary algorithms t...
Read article →Two groundbreaking preprints published today on arXiv herald a significant leap in artificial intelligence's ability to engage with the physical world, moving beyond isolated tasks to intricate collaborative activities and nuanced humanoid control. These papers, arXiv:2509....
Read article →A remarkable collection of eleven new research papers, all published or updated on arXiv CS. LG today, demonstrates a profound acceleration in how artificial intelligence is being integrated into fundamental scientific research and complex data analysis....
Read article →A wave of groundbreaking research, captured in recent arXiv preprints, reveals AI's accelerating trajectory into the most complex and computationally intensive corners of science and engineering. These papers, all published today, April 13, 2026, highlight a critical evolution: f...
Read article →The public text record, the very foundation upon which both humans and large language models (LLMs) learn, is increasingly being shaped by the LLMs’ own outputs. This recursive process, where generated text re-enters the training corpus, poses a fascinating and critical challenge...
Read article →A new wave of research, published today on arXiv CS. AI, tackles some of the most pressing challenges facing Large Language Models (LLMs): computational efficiency, robust reasoning, and the persistent problem of hallucination....
Read article →A new research paper has revealed the existence of 'correctness bugs' within torch. compile, the PyTorch compiler crucial for optimizing deep learning models, including large language models (LLMs)....
Read article →A flurry of groundbreaking research papers, all recently published on arXiv, signals a significant push in the AI community to enhance the reasoning and problem-solving capabilities of large language models (LLMs). These studies collectively address critical challenges from compu...
Read article →Microsoft is fundamentally overhauling its Windows Insider Program (WIP), making it significantly easier for testers to access cutting-edge experimental features without relying on third-party tools. This strategic simplification aims to make the beta program more predictable and...
Read article →