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 401–420
A flurry of groundbreaking research papers, all published today, April 21, 2026, on arXiv CS. LG, signals a significant acceleration in how artificial intelligence is being harnessed to tackle some of science and engineering's most computationally intensive and data-constrained c...
Read article →A crucial step toward more accessible and cost-effective artificial intelligence has been announced today with new research allowing developers to predict the performance degradation of large language models (LLMs) when compressed. This breakthrough promises to save significant c...
Read article →A new wave of research papers, published today on arXiv CS. LG, signals significant advancements across several critical areas of machine learning, from enhancing the training efficiency of large language models to enabling more robust scientific discovery and tackling real-world...
Read article →A flurry of eight significant research papers, all published today on arXiv, signal a concerted scientific effort to advance the core capabilities, reliability, and interpretability of artificial intelligence, particularly foundation models and agentic systems. These simultaneous...
Read article →Today's arXiv pre-prints unveil a torrent of innovation across AI, highlighting significant advancements in making large language models (LLMs) more efficient and trustworthy, even as new frameworks push the boundaries of AI-driven scientific discovery in fields like protein engi...
Read article →The latest wave of research papers from arXiv unveils significant advancements poised to democratize access to powerful AI models, enhance their safety, and deepen our fundamental understanding of their inner workings. Key among these are novel quantization techniques that dramat...
Read article →The bleeding edge of AI research has once again delivered a torrent of new insights, with a collection of papers appearing on arXiv today, all pointing towards a future of more efficient, robust, and controllable large language models (LLMs). From tackling the quadratic computati...
Read article →A significant new vulnerability, dubbed "LLM Hypnosis," has been identified, demonstrating that a single malicious user can persistently alter the knowledge and behavior of large language models (LLMs) through targeted prompts and feedback. This discovery, detailed in a recent ar...
Read article →A significant leadership transition is underway at Apple, as CEO Tim Cook is set to step down from his role, with John Ternus, currently Apple's senior vice president of hardware engineering, poised to take the helm. This pivotal change, announced on April 20th, 2026, will see Te...
Read article →The landscape of digital creativity is experiencing a profound transformation, with generative AI simultaneously ushering in new frontiers of interaction in gaming and posing significant challenges for music distribution. Both possibilities and pitfalls are unfolding rapidly, as...
Read article →Two distinct yet equally pivotal advancements have just emerged from the arXiv pre-print server, signaling a leap forward in both robot autonomy and human-computer interaction. Researchers have unveiled VeriGraph, a novel framework designed to enhance the reliability of robot tas...
Read article →Today, the arXiv preprint server has unveiled a quartet of pivotal research papers, all published on April 20, 2026, collectively advancing the theoretical underpinnings and architectural innovations across critical domains of artificial intelligence. These new works address chal...
Read article →A recent surge of research, all published on April 20, 2026, marks a significant inflection point for Large Language Model (LLM) agents, showcasing their growing sophistication in tool use, complex task execution, and real-world deployment. These five new arXiv papers collectivel...
Read article →The frontier of large language model (LLM) research is buzzing with breakthroughs, as evidenced by a wave of new papers published today on arXiv, tackling everything from boosting inference speed and enhancing safety to pioneering applications in drug design and robotics. These d...
Read article →A new paradigm in scientific collaboration is emerging with the introduction of CoLabScience, a proactive Large Language Model (LLM) assistant designed to autonomously engage and accelerate biomedical discovery arXiv CS. AI....
Read article →Reinforcement learning (RL) is rapidly advancing beyond theoretical simulations, with new research demonstrating its capability to manage live production software environments, secure networked systems, and optimize critical infrastructure. This marks a significant pivot from con...
Read article →The world of Large Language Models (LLMs) is buzzing with a fresh wave of innovation, as researchers unveil breakthroughs that tackle fundamental challenges in efficiency, reliability, and application. Among the most exciting developments are new architectures designed to manage ...
Read article →The nascent but rapidly evolving field of fusion energy, long a beacon of limitless clean power, is beginning to experience early signs of financial strain. Recent reports indicate that while fundamental research continues to advance, the investment landscape for fusion startups ...
Read article →A significant stride in reusable rocketry has been achieved with the successful test-firing of Starship V3, signaling rapid advancements in space launch capabilities. Concurrently, an Artemis II pilot has offered a compelling first-hand account of what it was truly like to fly an...
Read article →Netflix is set to integrate a TikTok-like vertical video feed within its apps this month, signaling a significant evolution in its content strategy. Crucially, the streaming giant also plans to broadly utilize artificial intelligence for both content recommendations and, notably,...
Read article →