AI & Machine Learning
244 articles RSS
Thomson Reuters Launches Thomson, an In-House AI Model Built on Reworked Qwen Weights, to Cut Anthropic Reliance
Thomson Reuters spent $40 million building an in-house legal AI model on a reworked Alibaba Qwen base, aiming to reduce dependence on Anthropic and other outside AI labs.
Anthropic Previews a Model Hardware Standard Letting AI Agents Directly Operate Lab Robots and Manufacturing Equipment
Anthropic opened a research preview of the Model Hardware Standard, letting AI agents read and write to lab and factory hardware through standardized drivers, cutting integration from weeks to hours in early tests.
AWS and Nvidia to Deploy 2 Million More GPUs in 2027-2028 as Demand Outpaces Prior 1-Million Commitment
AWS and Nvidia announced an expanded partnership to deploy 2 million additional GPUs across AWS data centers in 2027-2028, alongside new Vera CPU, networking, and robotics integrations.
FreeToken Lets Frontier Mixture-of-Experts Models Run on Consumer GPUs With Dynamic CPU-GPU Co-Execution
UC Berkeley and MIT researchers, with Databricks co-founders Matei Zaharia and Ion Stoica, released an open-source engine that serves up to 753-billion-parameter MoE models on a single workstation GPU.
MCR-Bench Study Finds Leading LLMs' Code-Review Accuracy Collapses as Review Rounds Pile Up
A new benchmark testing seven LLMs on real multi-round GitHub code reviews finds accuracy drops sharply as review rounds increase, exposing weak memory across rounds.
AWS Open-Sources Kiro Crew for Asynchronous AI Coding Agents
AWS open-sourced Kiro Crew, an agent-orchestration tool already used internally at Amazon by more than 39,000 developers, under an Apache 2.0 license.
Meta and UIUC Researchers Get an 8-Billion-Parameter Model to Match Claude Opus 4.5 With a Smarter Agent Harness
EvoHarness-RL trains Qwen3-8B to hit 96.9% on ALFWorld, edging out Claude Opus 4.5's 96.4% baseline by rethinking agent memory and state.
New SWE-Prime Method Trains AI Coding Agents on Just 10% of Data, Lifting Bug-Fix Accuracy Up to 24.2%
A new study finds that curating just 10% of AI coding-agent training trajectories outperforms training on the full dataset, with gains up to 24.2%.
Z.ai Launches GLM-5.3-Flash, an MIT-Licensed Coding Model That Was Secretly Topping Leaderboards as 'Ox-Alpha'
Z.ai released GLM-5.3-Flash under an MIT license, confirming it was the anonymous 'ox-alpha' model that had been topping OpenRouter and OpenCode leaderboards.
REFINE Multi-Agent LLM System Cuts Java Code Smells Up to 73% but Flags Assertion and Method-Removal Risks
A new preprint tests a multi-agent LLM refactoring pipeline on 450 Java files, cutting code smells up to 73% while flagging that assertions and public methods are sometimes silently removed.
Diagrid Catalyst 2.0 Brings Cryptographically Verifiable, Durable Execution to Ten AI Agent Frameworks
Diagrid's Catalyst 2.0 adds Dapr-based cryptographic attestation and automatic failure recovery for AI agents across ten frameworks, including LangGraph and Microsoft Agent Framework.
CNCF Graduates Kubeflow, Certifying the Kubernetes-Native AI Platform as Production-Ready
The Cloud Native Computing Foundation gave Kubeflow its top maturity tier, citing 6,600-plus contributors and 260 million PyPI downloads.