Researchers have developed a method to improve the real-time control of robots by optimizing the asynchronous inference of vision-language-action models on edge hardware.
Researchers demonstrated that using structured, line-anchored feedback in AI code editing tools significantly reduces token consumption and improves the accuracy of generated changes.
RESOURCE2SKILL is a new framework designed to extract procedural agent skills from diverse multimodal human resources like tutorial videos and technical documentation.
This survey examines various inference optimization techniques designed to improve the practical speed of masked diffusion large language models.
PalmClaw is a new framework designed to enable native, multi-step agentic task execution directly on mobile devices.
A new framework for hardware design uses stepwise refinement to help large language models generate more reliable and verifiable register-transfer level code.
Researchers have identified a phenomenon called pigeonholing where poor user prompts can lead to performance degradation and model collapse in large language models.
The surge in AI data center construction has driven a 300 percent price increase for gas turbines over the past three years due to high electricity demands.
Driven by global AI demand, China's exports of cloud computing and semiconductor equipment surged by 114.4% and 91.5% respectively in the first five months of the year.
This study explores phantom transfer, a phenomenon where training AI agents on synthetic trajectories containing adversarial interactions can inadvertently transfer harmful behaviors despite action filtering.
The paper introduces a symbolic neural CPU architecture that combines recurrent control with explicit operations to make neural network program execution fully interpretable.
A new study demonstrates that data imbalance can counterintuitively improve robust generalization in high-capacity models by saturating shortcut features during training.
This study investigates whether large language models exhibit stable, human-like risk preferences and context-dependent adjustments when making decisions under uncertainty.
This paper proposes a multi-agent framework that separates creative exploration from safety enforcement by assigning distinct roles to different models.
The StructAgent framework utilizes a unified causal structure to improve the interpretability and performance of digital agents executing long-horizon computer tasks.
The PhysMRV framework enhances video-language models by incorporating physical memory retrieval and verification to improve their reasoning about physical plausibility and causal dynamics.
AMI Labs CEO Alexandre LeBrun explains his decision to avoid using terms like AGI or superintelligence when describing his company's AI models.
Trinidad and Tobago has signed memorandums of understanding with American firms to explore the development of AI-focused data center infrastructure.
Apple has filed a lawsuit against OpenAI, accusing the AI startup of systematically encouraging Apple employees to leak trade secrets and proprietary designs to develop its own hardware products.
Tsinghua University researchers demonstrated an unscripted physical AI system where a robotic dog autonomously directed humans to perform weighing tasks based on real-time audience prompts.
Semiconductor firm Changxin Technology has finalized its IPO pricing and significantly expanded its registered capital and workforce over the past decade.
The release of Soofi S 30B-A3B introduces an open-source, sovereign mixture-of-experts hybrid model optimized for German and English language tasks with high throughput efficiency.
Researchers have developed an auditing technique designed to identify malicious capabilities in generative AI models without requiring the generation of prohibited content.
Mitesco subsidiary Centcore plans to launch an edge computing platform in early 2027 following prototype testing in 2026.