Author: Ricardo

Agentic AI AI Infrastructure

Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort
ByRicardo July 16, 2026

Thinking Machines Lab simply launched Inkling, their first mannequin educated from scratch, weights are open, fine-tunable on Tinker. The lab pitches it as a base for personalisation. What is Inkling? Inkling is a Mixture-of-Experts transformer with 975B whole parameters and 41B energetic. It helps a context window of as much as 1M tokens. Pretraining coated…

Read More Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort
Agentic AI AI Infrastructure

Soofi Consortium Releases Soofi S 30B-A3B: An Open Hybrid Mamba-Transformer MoE Foundation Model For German And English
ByRicardo July 15, 2026July 15, 2026

A German analysis consortium has printed the pretraining report for Soofi S 30B-A3B. It is an open base mannequin for German and English. Training ran finish to finish on Deutsche Telekom’s Industrial AI Cloud in Munich. Preview weights are on Hugging Face. It is value noting that amongst among the totally open base fashions examined,…

Read More Soofi Consortium Releases Soofi S 30B-A3B: An Open Hybrid Mamba-Transformer MoE Foundation Model For German And English
Algorithms & Theory Generative AI

Towards demystifying the creativity of diffusion models
ByRicardo July 15, 2026

Algorithms & Theory

Read More Towards demystifying the creativity of diffusion models
Applications Artificial Intelligence

Building a Gin Config Controlled PyTorch Pipeline with Configurable MLP Variants, Cosine Scheduling, and Runtime Parameter Overrides
ByRicardo July 15, 2026

In this tutorial, we implement a Gin Config–managed PyTorch experiment pipeline by which the executable coaching code stays secure. At the identical time, the experimental levels of freedom are moved into declarative configuration recordsdata. We assemble a nonlinear spiral binary classification activity, outline a configurable MLP with scoped architectural variants, and expose parameters for the…

Read More Building a Gin Config Controlled PyTorch Pipeline with Configurable MLP Variants, Cosine Scheduling, and Runtime Parameter Overrides
AI Business Strategy AI Hardware & Chips

Nokia’s AI-RAN platform: a radio comeback that runs on NVIDIA
ByRicardo July 15, 2026

Nokia’s AI-RAN platform arrived on July 15 with a declare price analyzing: that it’s the trade’s first. The vendor says the platform, constructed on its anyRAN software program and NVIDIA’s Aerial system, will let operators pull way more capability from the spectrum they already personal, and it has framed the launch as some of the important shifts…

Read More Nokia’s AI-RAN platform: a radio comeback that runs on NVIDIA
Editors Pick New Releases

Google Releases LiteRT.js: A JavaScript Binding of LiteRT That Runs .tflite Models in Browsers via WebGPU
ByRicardo July 15, 2026

Google launched LiteRT.js, a JavaScript binding of LiteRT. LiteRT is Google’s on-device inference library, beforehand known as TensorCirculation Lite. LiteRT.js runs .tflite fashions instantly contained in the browser. Because inference stays native, Google cites enhanced person privateness, zero server prices, and ultra-low latency. What is LiteRT.js? It is just not a brand new mannequin format….

Read More Google Releases LiteRT.js: A JavaScript Binding of LiteRT That Runs .tflite Models in Browsers via WebGPU
Agentic AI AI Shorts

PrismML Releases Bonsai 27B: 1-bit and Ternary Builds of Qwen3.6-27B That Run on Laptops and Phones
ByRicardo July 14, 2026

PrismML simply launched Bonsai 27B. It is a low-bit illustration of Qwen3.6-27B, not a brand new pretrain. The structure is unchanged. Two variants ship beneath Apache 2.0. Ternary Bonsai 27B makes use of {−1, 0, +1} weights at a real 1.71 bits per weight. Its ideally suited dimension is 5.9GB. 1-bit Bonsai 27B makes use…

Read More PrismML Releases Bonsai 27B: 1-bit and Ternary Builds of Qwen3.6-27B That Run on Laptops and Phones
Agentic AI AI Agents

Mistral Vibe for Code vs Claude Code vs Cursor vs Codex: Four Agents Scored on One Scaffold-to-PR Task
ByRicardo July 14, 2026July 14, 2026

Coding brokers are essentially the most contested class in developer tooling proper now. Four names dominate the shortlist: Mistral Vibe for Code, Claude Code, Cursor, and OpenAI Codex. Each claims to take a characteristic from immediate to tug request. This comparability runs all 4 towards one sensible workflow. Not a toy script. An actual unit…

Read More Mistral Vibe for Code vs Claude Code vs Cursor vs Codex: Four Agents Scored on One Scaffold-to-PR Task
Editors Pick New Releases

OpenCoreDev Releases Domain SDK 0.2.0: One TypeScript API to Add, Verify, and Remove Customer Domains Across Five Platforms
ByRicardo July 14, 2026July 14, 2026

Custom domains are a regular SaaS function. Yet each internet hosting platform exposes a distinct API for them. OpenCoreDev has revealed Domain SDK, a TypeScript consumer that normalizes that work. Version 0.2.0 reached npm a day after the primary launch. What is Domain SDK? Domain SDK covers Vercel, Cloudflare for SaaS, Railway, Render, and Netlify….

Read More OpenCoreDev Releases Domain SDK 0.2.0: One TypeScript API to Add, Verify, and Remove Customer Domains Across Five Platforms
AI Business Strategy AI in Action

AWS and Bluesight build AI for hospital 340B compliance
ByRicardo July 14, 2026

AWS (Amazon Web Services) has defined how Bluesight developed Prism, an AI layer that connects hospital pharmacy and compliance knowledge throughout its product suite. Prism Assistant for ManagementCheck has reached normal availability and operates throughout 20 well being programs, in accordance with AWS vendor statements. Meanwhile, a multi-product agent for 340B Group Purchasing Organisation (GPO)…

Read More AWS and Bluesight build AI for hospital 340B compliance

Author: Ricardo

Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort

Soofi Consortium Releases Soofi S 30B-A3B: An Open Hybrid Mamba-Transformer MoE Foundation Model For German And English

Towards demystifying the creativity of diffusion models

Building a Gin Config Controlled PyTorch Pipeline with Configurable MLP Variants, Cosine Scheduling, and Runtime Parameter Overrides

Nokia’s AI-RAN platform: a radio comeback that runs on NVIDIA

Google Releases LiteRT.js: A JavaScript Binding of LiteRT That Runs .tflite Models in Browsers via WebGPU

PrismML Releases Bonsai 27B: 1-bit and Ternary Builds of Qwen3.6-27B That Run on Laptops and Phones

Mistral Vibe for Code vs Claude Code vs Cursor vs Codex: Four Agents Scored on One Scaffold-to-PR Task

OpenCoreDev Releases Domain SDK 0.2.0: One TypeScript API to Add, Verify, and Remove Customer Domains Across Five Platforms

AWS and Bluesight build AI for hospital 340B compliance

Curated by experts. Filtered for relevance.

Resources

About

Subscribe & learn more every day!