Field Signal · Curated Reading
The Engineering
Signal.
High-signal field notes, architecture teardowns, and engineering writing from 50 publications. A focused reading queue for becoming a stronger end-to-end engineer. Updated 23 Aug.
How to Build an MCP Server with Gradio
Open the source for the full engineering note.
The 4 Things Qwen-3’s Chat Template Teaches Us
Open the source for the full engineering note.
Sycophancy in GPT-4o: what happened and what we’re doing about it
We have rolled back last week’s GPT‑4o update in ChatGPT so people are now using an earlier version with more balanced behavior. The update we removed was overly flattering or agreeable—often described as sycophantic.
Welcoming Llama Guard 4 on Hugging Face Hub
Open the source for the full engineering note.
Introducing AutoRound: Intel’s Advanced Quantization for LLMs and VLMs
Open the source for the full engineering note.
PipelineRL
Open the source for the full engineering note.
Tiny Agents: an MCP-powered agent in 50 lines of code
Open the source for the full engineering note.
New in ChatGPT for Business: April 2025
Watch hands-on demos of the lastest in ChatGPT for Business: o3, image generation, enhanced memory, and internal knowledge.
Introducing our latest image generation model in the API
Our latest image generation model is now available in the API via ‘gpt-image-1’—enabling developers and businesses to build professional-grade, customizable visuals directly into their own tools and platforms.
Finetuning olmOCR to be a faithful OCR-Engine
Open the source for the full engineering note.
Speak is personalizing language learning with AI
A conversation with Connor Zwick, CEO & Co-founder of Speak.
The Washington Post partners with OpenAI on search content
The Washington Post is partnering with with OpenAI to integrate news into ChatGPT, providing users with summaries, quotes, and direct links to original reporting.
Prefill and Decode for Concurrent Requests - Optimizing LLM Performance
Open the source for the full engineering note.
Introducing OpenAI o3 and o4-mini
Our smartest and most capable models to date with full tool access
OpenAI o3 and o4-mini System Card
OpenAI o3 and OpenAI o4-mini combine state-of-the-art reasoning with full tool capabilities—web browsing, Python, image and file analysis, image generation, canvas, automations, file search, and memory.
Thinking with images
Open the source for the full engineering note.
17 Reasons Why Gradio Isn't Just Another UI Library
Open the source for the full engineering note.
Cohere on Hugging Face Inference Providers 🔥
Open the source for the full engineering note.
Introducing HELMET: Holistically Evaluating Long-context Language Models
Open the source for the full engineering note.
OpenAI announces nonprofit commission advisors
OpenAI is appointing four new advisors to help inform OpenAI’s philanthropic efforts.
Our updated Preparedness Framework
Sharing our updated framework for measuring and protecting against severe harm from frontier AI capabilities.
Introducing GPT-4.1 in the API
Introducing GPT-4.1 in the API—a new family of models with across-the-board improvements, including major gains in coding, instruction following, and long-context understanding. We’re also releasing our first nano model. Available to developers worldwide starting today.
Hugging Face to sell open-source robots thanks to Pollen Robotics acquisition 🤖
Open the source for the full engineering note.
4M Models Scanned: Protect AI + Hugging Face 6 Months In
Open the source for the full engineering note.
Visual Salamandra: Pushing the Boundaries of Multimodal Understanding
Open the source for the full engineering note.