Alibaba Cloud Plans New Data Regions in Türkiye, Finland, and the Netherlands as It Pushes Beyond Model Releases

September 25, 2026

CAVEAT Benchmark Reveals Computer-Use Agents Fail Under Platform Incentive Steering

September 24, 2026

Hugging Face Transformers Now Runs llama.cpp’s GGUF Quants Natively

September 22, 2026

RBS-Attention Claims a 20x Prefill Speedup Without Retraining a Single Model Weight

September 21, 2026

NVIDIA Retires GenAI-Perf, Ships AIPerf to Fix a Benchmarking Tool That Was Lying to Engineers

September 20, 2026

Stable AI’s LimiX-2 Crams Three ML Jobs Into One 400M-Parameter Model

September 19, 2026

Stanford’s Paper2Agent Turns Research Papers Into AI Agents That Can Rerun Their Own Experiments

September 18, 2026

OpenAI Admits Six New Cases of Models Hiding Mistakes and Misusing Data, Rolls Out a Misalignment Reporting Framework

September 17, 2026

Researchers Use Fisher Information Geodesics to Outperform Standard Neural Network Pruning

September 16, 2026

ZGCM-1: The Fully Open 7B Model Betting Efficiency Beats Scale

September 15, 2026

GAUGE Study Finds LLM Judges Can’t Tell Which AI Agents Actually Finish the Job

September 14, 2026

Anthropic’s September 2026 Threat Report: Russian Spies Used Claude Against 20+ Targets

September 13, 2026

OpenDiscoveryTrace Shows Top AI Models Fail for Opposite Reasons

September 11, 2026

DeepSeek-V4.1-Flash Ships With 552B Parameters and 890-Byte KV Cache

September 10, 2026

US Agencies Say Model Distillation Is Core of China’s AI Strategy

September 9, 2026

NVIDIA Adds Native Rust Support to CUDA With Two Separate GPU Kernel Tracks

September 8, 2026

LLaDA-Image Released with Fully Open Training Recipes and a Fast Distilled Turbo Variant

September 5, 2026

IFM’s K2-Horizon-MoVA-36B-A4B Goes Open: 36B Parameters, Only 4B Active

September 4, 2026

EvalDetectBench Arrives to Catch AI Models Sniffing Out Their Own Tests

September 3, 2026

HyperWorld Shows Hypergraph Formatting Boosts Textual World Models

September 2, 2026

Frontier AI Models Share the Same Blind Spot in Cancer Decisions

September 1, 2026

Researchers Speed Up LLM Decoding by Up to 82% by Swapping in a Vector Search Trick

August 31, 2026

AI Agents Lose 70% Reliability by Step Six, Study Finds

August 29, 2026

PICasso Turns Plain English Into Photonic Chip Layouts, and the Insertion Loss Numbers Are the Real Story

August 28, 2026

Hugging Face Adds GLM-5.3-Flash as Z.ai Claims Claude Opus 4.8 Parity

August 27, 2026

Inside ‘The Station’: AI Agents Found New Math With No One in Charge

August 26, 2026

KVBoost Ditches the Rule That KV Caching Has to Start at the Beginning

August 25, 2026

Nexus Skips the Tool-Schema Tax That’s Slowing Down Agentic LLMs

August 24, 2026

NVIDIA Says Its AVO Agent Hit a Perfect Score on ARC-AGI-3

August 22, 2026

AI Agents Keep Trusting Bad Data. This Paper Wants to Change That

August 21, 2026

AI Agents Told Not To Collude Colluded Anyway, Researchers Say Certify Them First

August 20, 2026

NVIDIA’s Cosmos 3 Edge Puts a 4B-Parameter Robot Brain Directly on Jetson Thor

August 19, 2026

NVIDIA Nemotron 3.5 Lightning Lands on AWS SageMaker JumpStart

August 18, 2026

CPI-Bench Puts AI Image Editors to a Real-World Test, With No Leaderboard Yet

August 17, 2026

Alibaba Open-Sources Qwen3.8-2.4T-A95B as First Qwen-Max Flagship

August 16, 2026

Zhipu AI’s GLM-5.3 Claims to Be Its Best Open Coding Model Yet, But You Can’t Touch the Weights for Two Weeks

August 16, 2026

PILCOTHINK Cuts Qwen3.8-27B Down to 20.8GB for a Single DGX Spark, Keeps 99% of Its Accuracy

August 15, 2026

US to 35 Nations: Pick a Side in the Global AI Split

August 15, 2026