A new recipe distills Pythia-1B into a Mamba model, achieving a perplexity of 14.01 compared to the teacher's 13.86.
The 321B parameter model based on GLM-5.3-Flash defaults to max reasoning, requiring token usage checks before comparison.
The connector allows Claude to access Playwright traces, network requests, and flakiness history to diagnose CI test failures accurately.
The agent reached 10 million U.S. App Store downloads within 40 days of launch, joining only 11 other apps to hit that milestone.
A guide explains how content creators can use ElevenLabs' voice cloning tools, including Eleven v4 and ElevenCreative, to produce multilingual and consistent content.
The platform offers curated industry-specific AI agents for finance, healthcare, and retail, with purchasing via ElevenLabs commitments planned for next year.
The program targets $3 in services revenue for every $1 in ElevenLabs revenue, supported by 1,100 certified deployment engineers across 350 partner organizations and 550+ active partners.
The alliance combines McKinsey's QuantumBlack AI engineering with ElevenLabs' ElevenAgents platform to help clients prove value and scale deployments across industries like insurance and telecom.
KPN fine-tuned models to raise Dutch identity-verification accuracy from 50% to 90%, while Yahoo Scout exited beta after 3 billion answers.
The company introduced new agent features including Memory and Routines, expanded its partner network to 550 firms, and launched a marketplace for industry-specific AI applications.
The partnership includes a sovereign deployment for Poland's e-Health platform and plans to add 15MW of NVIDIA Vera Rubin capacity.
The service allows teams to run the entire agent control plane, including harness, sandboxes, and credentials, in their own cloud or on-premises environment.
Anthropic has released Claude Fable 5.2 and Mythos 5.2, lowering prices and improving output speed by up to 30%.
The desktop app allows coding agents to open a focused Unity editor with the real Inspector for specific objects and supports up to three multiplayer clones.
InfraRSI achieved a geometric mean speedup of 1.83x on H100/H200 by simultaneously exploring configurations and code.
Security researchers identified 2,539 real secrets exposed in public GitHub repositories through the .claude/settings.local.json file.
The model supports 43 languages and runs on phones, laptops, and browsers using Apache 2.0 licensed weights.
The model couples a Qwen3.5-9B language model with a 29B MMDiT diffusion backbone to predict video outcomes of actions.
AI Briefing