Latest News
Depth Anything V2 Boosts Accuracy in Robotic Surgery Depth Perception
New research uses synthetic data and DV-LORA to improve monocular depth estimation, a key factor for precision in complex surgical environments. Achieves top results on the SCARED dataset.
DreamTacVLA Gives Robots a High-Res Sense of Touch
A new framework adds detailed tactile sensing to VLA models, helping robots master contact-heavy tasks with 95% accuracy.
GR-Dexter Advances Bimanual Dexterous-Hand Robot Manipulation
New framework combines custom hardware, intuitive teleoperation, and curated datasets to boost vision-language-action models for bimanual robots.
DiffThinker: Redefining Multimodal AI with Generative Image Reasoning
DiffThinker introduces a new diffusion-based method that treats multimodal reasoning as a generative image-to-image task, surpassing top models in vision-focused challenges.
OpenScholar Beats GPT-4o in Scientific Research Synthesis
An open-source model from UW and AI2 outperforms GPT-4o in summarizing research and citing sources, earning scientists’ preference.
TTT-Discover: AI Trains During Use, Beats Human Experts
A new method lets AI keep learning while running, finding fresh solutions and outperforming humans on tough tasks.
Why AI Projects Are Failing: Over-Optimism, Data Flaws, and the Rise of DAG Workflows
From runaway chatbots to data pitfalls, the AI industry pivots to controlled workflows and stronger data governance—highlighted by the Air Canada chatbot fiasco.
New Method Maps Entropy in Phase-Change Materials to Boost Memory Speed
Researchers combine first-principles calculations, machine learning, and experiments to track entropy in phase-change materials, aiming for faster, more efficient memory devices.
Chinese Team Uses Quantum Entanglement to Block Hardware Backdoors
A breakthrough in quantum data transmission by a Chinese research team could end hardware-based cyberattacks. Is this the next frontier in secure communication?
DHS Uses AI Video Generators, Raising Ethics and Transparency Concerns
The Department of Homeland Security is using AI video tools from Google and Adobe, sparking debate over ethics, bias, and echoes of past immigration policies.
Plasmon Computing Could Slash AI’s Power Demand
Hector De Los Santos and his team are pioneering plasmon computing—using electron waves instead of current—to tackle AI’s growing energy problem.
DeepRead: RAG Agent Uses Document Structure to Deliver Smarter Answers
Researchers unveil DeepRead, a new Retrieval-Augmented Generation (RAG) agent that reads document structure to boost question-answering accuracy.
RaBiT Sets New Standard for 2-Bit LLM Quantization
RaBiT, a new quantization method, solves feature co-adaptation in residual binarization, delivering state-of-the-art 2-bit performance and major inference speed gains for large language models.
OpenAI and Snowflake Join Forces to Embed AI in Enterprise Data
A $200M deal will integrate OpenAI’s AI models directly into Snowflake’s platform, reshaping how businesses use their data.
New AI Model Integrates Multimodal Data to Improve Ameloblastoma Diagnosis
Researchers introduce an AI model and dataset combining radiology, pathology, and clinical data to enhance diagnosis and surgical planning for ameloblastoma.