News
Nvidia to Back $105 Billion OpenAI Data Center in Ohio
Nvidia agreed to guarantee up to $105 billion in lease and power payment obligations to support a new data center…
News
Anthropic Raises Misalignment Risk Rating to ‘Low’, Shelves Internal Model 2
Anthropic published its August 2026 Risk Report, raising its assessment of catastrophic misalignment risk in high-stakes settings from “very low”…
News
Anthropic’s Annualized Revenue Surges to $65 Billion Ahead of Planned IPO
Anthropic’s annualized revenue run rate reached $65 billion by late July 2026, up sharply from $47 billion in May and…
Engineering
Ventor-QTest: Threat-Model-Driven Verification of Vendor-Hosted LLM APIs
Ventor-QTest is a black-box auditing framework for assessing the inference quality of vendor-hosted large language model APIs without access to…
Engineering
Improving the Matrix Multiplication Exponent with Modern Optimization and AlphaEvolve
Researchers improved the theoretical upper bound on the matrix multiplication exponent to 2.371177, down from the prior best of 2.371339,…
Engineering
DumpsterCluster: From Dumpster Diving to Serving LLaMA-70B on $60 GPUs
Researchers at the University of Oxford built a 128-GPU cluster from second-hand NVIDIA V100 processors, costing about $22,000, and used…
Engineering
When Context Bites: Detecting RAG Poisoning via Document-Level Attention Collapse
Researchers identified a vulnerability in retrieval-augmented generation systems in which adversarial document injection causes models to answer with false confidence…
Engineering
Agentic Transaction: Towards ACID-Compliant Agent Systems
Researchers from Tsinghua University proposed an ACID-compliant framework that adapts classical database transaction principles to autonomous LLM agent execution, introducing…
Engineering
Learn What’s Left, Not What’s Mastered: Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization
Researchers introduced Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization (SA-MRPO), a method for multi-objective reinforcement learning that normalizes each…
Engineering
How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research Tasks
Researchers built AutoResearchEval, a benchmark of 100 frontier research tasks spanning seven scientific domains and the full research lifecycle, and…
News
DeepSeek Ships V4-Pro Flagship Model With Trillion-Parameter Scale and Agent Focus
DeepSeek released its flagship V4-Pro model (build 0813) on August 12, 2026, ending a preview period that began earlier in…
News
Apple Trained Its Own AI Model for China With Alibaba’s Help, Wins Beijing Clearance
Apple has developed a proprietary large language model for the Chinese market with technical support from Alibaba, moving away from…