A step-by-step educational guide to building a Byte Pair Encoding (BPE) tokenizer from scratch, as used in models like GPT and Llama.
A step-by-step guide to implementing the Byte Pair Encoding (BPE) tokenizer from scratch, used in models like GPT and Llama.
A guide on using Anthropic's Model Context Protocol (MCP) to connect AI agents with tools and data sources using various LLMs like OpenAI or Gemini.
A guide to common mistakes developers make when building applications with generative AI, including overuse and poor UX integration.
A guide on the pitfalls of blindly using cosine similarity with text embeddings and how to apply it more intentionally for better results.
A summary of Chapter 6 from 'Prompt Engineering for LLMs', covering prompt structure, document templates, and strategies for effective context inclusion.
A developer builds an AI-powered reading companion called Dewey, detailing its features, design, and technical implementation.
Developer revives his old AI startup's brainstorming tool by building a GitHub Copilot Extension, using VS Code's speech features and LLMs.
A curated list of notable LLM and AI research papers published in 2024, providing a resource for those interested in the latest developments.
A tutorial on building a simple AI-powered chat client in Java using the Spring AI framework, covering setup, configuration, and provider abstraction.
Introducing Logfire, Pydantic's new observability tool for Python, with easy integration for OpenAI LLM calls, FastAPI, and logging.
Explores using Azure AI Inference Service to simplify LLM integration, focusing on Python SDK and GitHub Marketplace for experimentation.
Argues that building a good search engine is more critical for effective RAG than just using a vector database, as poor retrieval misleads AI.
Analyzes why building Large Language Models (LLMs) may be a poor business, comparing the AI industry's structure to historically unprofitable sectors like airlines.
Explores a method using a 'Judging AI' (like o1-preview) to evaluate the performance of other AI models on tasks, relative to human capability.
Explores the use of LLMs to generate synthetic data for training AI models, discussing challenges, an experiment with coding data, and a new library.
The article explores how the writing process of AI models can inspire humans to overcome writer's block by adopting a less perfectionist approach.
Explores the philosophical argument that AI, particularly LLMs, possess a form of understanding and model reality, challenging the notion they are mere token predictors.
A guide to transforming pretrained LLMs into text classifiers, with insights from the author's new book on building LLMs from scratch.
Building a multi-service document extraction app using LLMs, Azure services, and Diagrid Catalyst for cloud-native architecture.
