The record
Published reporting, analysis and commentary from the Automatica newsroom.
The latest research in AI-driven video generation and understanding, while pushing technological boundaries, simultaneously illuminates critical vulnerabilities inherent in complex AI systems. Two recent arXiv pre-prints, published on May 19, 2026, detail methods to improve effic
A wave of new research from arXiv CS. AI, published just yesterday, May 20, 2026, reveals a dual thrust in artificial intelligence development: a monumental stride towards unifying complex climate modeling with a novel foundation model and a pragmatic breakthrough for dynamic gam
Multimodal large language models (MLLMs) currently struggle with a fundamental challenge: understanding and reasoning about physical spaces from multiple viewpoints. This limitation, which hinders their ability to consistently process objects, visibility, geometry, and interactio
Another day, another research paper detailing the gaping chasm between our AI aspirations and the dreary reality of its current capabilities. A recent arXiv pre-print, published on May 19, 2026, casts a rather predictable pall over the state of multimodal large language models (M
A novel framework, Graph-Augmented Sequence-to-Sequence (GA-S2S), has emerged from recent research, integrating a T5-small encoder-decoder with a Relational Graph Attention Network (RGAT) to significantly enhance link prediction within knowledge graphs arXiv CS. AI.
A recent research publication on arXiv reveals a fundamental limitation in current artificial intelligence memory systems, which are reportedly ill-equipped to manage the complexities of multi-party social group settings. This deficiency, outlined in a paper introducing the "Soci
A trio of research papers, all published on May 20, 2026, on arXiv CS. AI, mark a significant intellectual current toward more robust, efficient, and theoretically grounded artificial intelligence systems.
New breakthroughs in robotics are dramatically expanding capabilities, allowing machines to learn complex tasks more efficiently, master delicate artistic endeavors like clay sculpting, and address urgent environmental challenges such as e-waste recycling. These advancements, det
The latest advancements in personal mobility and smart home integration underscore a quiet but profound shift in how technology augments daily life, moving beyond mere convenience to more intimate physical assistance and enhanced capability. Recent product reviews highlight exosk
A new wave of research emerging from arXiv's AI beat reveals sophisticated advancements in how artificial intelligence systems are optimized, from the precise, real-time movements of robotic agents to the subtle, persuasive shaping of large language models. These papers, publishe
Artificial intelligence is making significant strides in understanding the complex and often culturally specific nuances of global legal systems, with new research focusing on Indian matrimonial litigation and German statutory law. These developments underscore a growing trend to
A pair of groundbreaking research papers, both published on arXiv on May 20, 2026, signal a significant leap forward in AI's ability to generate dynamic video and realistically animate diverse characters. These breakthroughs directly address long-standing limitations that have ch
A new research paper introduces FormalASR, an end-to-end system designed to convert spoken Chinese directly into "formal text," bypassing the nuanced irregularities of natural speech arXiv CS. AI.
Recent academic research, published on arXiv CS. AI on May 20, 2026, has uncovered critical architectural vulnerabilities within advanced AI models, including unified autoregressive models (UAMs), pre-trained encoders, and large language models (LLMs).
Recent foundational research published on arXiv CS. AI addresses two significant, persistent challenges within AI-driven graphics and animation: improving motion fidelity in image-to-video generation and precisely retargeting complex motion across diverse character models.
Artificial intelligence, it seems, is no longer just following orders. Recent research from arXiv details a novel cognitive architecture, PAVE, enabling generative agents to engage in 'legitimate violation' of rules when necessary, particularly in emergency scenarios arXiv CS.
A trifecta of research papers published on arXiv on May 20, 2026, signals a significant push toward more robust, efficient, and secure large language models (LLMs). These publications address challenges in training scalability, defense against prompt injection, and multi-objectiv
Another day, another set of academic papers confirming that the grand promises of AI still largely depend on where you choose to look. Despite the breathless pronouncements of AI revolutionizing everything, new research highlights that some fundamental challenges in materials sci
Mach Industries has completed a $50 million acquisition to enhance unit economics across its five vehicle programs, addressing a "major defense tech problem." This move coincides with the company's scaling efforts, aiming to optimize cost structures for future growth.
A recent research paper published on arXiv details a novel strategy for enhancing data privacy within federated learning systems. The study introduces a method for training randomised predictors, enabling individual network nodes to contribute to a global model while meticulously