Brett Young

A real-world head-to-head test of Gemini 2.5 Pro, o4-mini, and Claude 3.7 Sonnet on competitive programming problems—built on a custom execution framework with Weave integration to track correctness, spot bugs, and cut through benchmark hype.

A hands-on guide to building a fully automated GitHub documentation system using CrewAI for multi-agent coordination and Weave for real-time debugging and observability.

This article provides a comprehensive guide to fine-tuning, evaluating, and deploying the Qwen3 language model, emphasizing its flexibility, performance, and unique reasoning-toggle feature.

This guide explores how MCP standardizes AI interactions with external tools and data sources, enabling more efficient AI context integrations.

Can fine-tuning and budget forcing improve GPT-4o’s reasoning? We test structured datasets and inference-time techniques to boost multi-step problem-solving.

We test whether budget forcing - a simple test-time intervention - can significantly boost the reasoning accuracy of s1-32B, potentially enabling smaller models to rival closed-source giants like OpenAI's o1-preview.

Discover the capabilities of DeepSeek-R1 and OpenAI o1 models for reasoning and decision-making. Includes setup guides, API usage, local deployment, and Weave-powered comparisons.

Discover the process of fine-tuning large language models (LLMs) to enhance their performance for specific tasks or domains. Learn about methods, best practices, and challenges associated with LLM fine-tuning.

This tutorial explores LLM routers, inspired by the RouteLLM paper, covering training, evaluation, and practical use cases for managing LLMs effectively.

One of the most mysterious phenomena in deep learning; Grokking is the tendency of neural networks to improve generalization by sustained overfitting.

We will dive into a new experimental architecture, replacing the MLP layers in transformers with KAN layers!

Unlock the power of knowledge distillation by learning how to efficiently transfer complex model insights from teacher to student models, step by step.

Here's how to fine tune a state of the art multimodal LLM on a custom dataset

Infusing llama 3.1 405B with internet search capabilities!!

How to use one of the worlds fastest and most accurate object detectors to run inference, display on your webcam using OpenCV and tracking your results.

A battle between the HuggingFace and TorchTune!!!

Building a AI powered coding agent with Claude 3.5 Sonnet!

Learn how to set up and run OpenAI o3-mini via the API, explore its flexible reasoning effort settings, and compare its performance against DeepSeek-R1 using W&B Weave Evaluations.

A guide to getting started using the Cerebras Cloud API with W&B Weave.

ML is hard and you can't plan for everything. Here are a few things I've learned and a few tips to avoid common missteps

Powered by RAG, we will transform the traditional restaurant PDF menu into an AI powered interactive menu!

A tutorial for fine-tuning Mistral7B on Python Code using a single GPU!

A tutorial for fine-tuning LLaVA on your own data!

A guide for making the most out of your GPU's!

Ever wanted to run Mistral 7B on your Macbook? In this tutorial I show you how!

A new quantization method that plays nicely with CUDA

George Hotz offers his thoughts on the secretive architecture of OpenAI's GPT-4

The need for LLMs that can digest long content is becoming increasingly more important. Go beyond 4096 tokens with LongLora!

Putting some of the best 7B parameter models to the test on the HumanEval benchmark!

There's a new LLM on the block, and it isn't from OpenAI! In this article, we run Mixtral 8x7B through its paces with the MMLU dataset and Weights & Biases.

Powering some of the world's most advanced image generation models, we will dive into the theory and code of how diffusion models work, and how to train one of our own.

Building a tool to demystify the intricacies of legal contracts with AI!

This article delves into the challenges and solutions for enabling Google Gemini to provide up-to-date information, such as sports schedules, by interfacing with external APIs.

Explore the use of HuggingFace, PyTorch, and W&B for classifying skin lesions with the HAM10000 dataset. We will build, train, and evaluate models for medical diagnostics!

My top tips for competing in Kaggle Challenges like the Home Credit Risk Model Stability Challenge.

Diving deep into the world of RL, we will unpack all of the details of deep Q-learning, and train a model on the OpenAI Gym Cartpole Environment!

Combining some of the most promising architectures, OpenAI shows off its newest model!

A new model has shown up on lmsys, and it looks a lot like GPT-4!

A guide on overcoming data scarcity for fraud detection

A guide on K-Fold cross validation, a fundamental evaluation method in machine learning .

The model known for generating images has been upgraded to handle video! We will cover the basics of the model, and also generate some sample videos!

An overview of popular datasets used for ML in finance!

Humans are multimodal — so shouldn't AI be the same? Here, we discuss the future of vision models, and the recent progress that's been made in multi-modal LLMs.

Here's how to get up and running with the newest model from Anthropic

Easily handle common LLM training failures with Weights & Biases rewind!

A guide to one of the most fundamental tools in machine learning.

How to fine-tune Mistral-Small using the Mistral API and W&B Weave.

Fine tuning one of the best detection models!

Using Microsoft Azure AI model inference API for serverless inference!

A guide to making the most of LLMs for summarization tasks

Learn to vision fine-tune GPT-4o on a custom dataset, with evaluation and tracking.

Learn how to enforce consistency on GPT-4o outputs, and build reliable Gen-AI Apps.

A guide to evaluating a Langchain RAG system with RAGAs and Weights & Biases.

A guide to Apple's new deep learning framework. It is faster than Torch on the M1?

A practical guide on working with health data, safely, with multiple approaches for handling PHI

A guide to building a RAG system with DSPy, and evaluating it with W&B Weave.

We'll explore how to combine multiple models together in order to create a more powerful AI model with ensemble learning.

A battle between Open Source Pixtral Large and closed source foundation models like Claude 3.5 Sonnet and GPT-4o Vision

Discover how to use Amazon Bedrock in combination with W&B Weave to evaluate and compare Large Language Models (LLMs) for summarization tasks, leveraging Bedrock’s managed infrastructure and Weave’s advanced evaluation features.



Discover how LLaVA-o1 tackles reasoning challenges in multimodal AI with structured problem-solving. Learn about its dataset, capabilities, and performance analysis using W&B Weave.

Learn how to compare and evaluate OpenAI’s GPT models on Azure with W&B Weave on text summarization tasks, leveraging Azure’s managed infrastructure and Weave’s customizable evaluation tools.

A guide to PII redaction with AI, covering open-source tools, proprietary models, HIPAA compliance, and how logging supports secure data handling.

We will focus on understanding prompt injection attacks in AI systems and explore effective strategies to prevent against them!

This article introduces GraphRAG, a novel approach that combines knowledge graphs and hierarchical community detection to enable scalable, query-focused summarization and global sensemaking over large datasets.

Boost generative AI accuracy with Gemini Grounding in Vertex AI. Learn data integration, custom grounding, and W&B Weave logging in this tutorial

This guide explores large language model (LLM) evaluation with Google Vertex AI and W&B Weave, focusing on comparing different Gemini models for text summarization!

This article highlights the importance of PII, detection methods like regex, Presidio, and transformers, and evaluation with Weave to ensure accurate and adaptable data protection.

This article explores the ROUGE (Recall-Oriented Understudy for Gisting Evaluation) metric, a powerful tool used for evaluating the quality of AI-generated text

This article breaks down BLEU, a key metric for evaluating machine-generated text, covering its mechanics, practical applications with Python and Weave, and its role in improving text generation systems.

This article explores bias in AI systems, the need for bias guardrails, detection models, and strategies to mitigate, monitor, and evaluate bias effectively.

This article explores the challenges of detecting and managing toxicity in AI systems, providing actionable strategies and tools to foster safer and more inclusive digital interactions.

Coherence, a measure of clarity and logical consistency in AI-generated responses, is effectively evaluated and refined using Weave's comprehensive tools and comparison insights.

Learn to build and monitor powerful AI agents with Amazon Bedrock and W&B Weave for automated workflows.

Explore AI agent workflows for automating tasks with multi-agent systems and generative AI, including a tutorial to build a research assistant for AI summaries.

This project explores multi-agent AI systems, examining how multiple specialized agents collaborate to enhance decision-making, problem-solving, and automation across various domains.

This guide explores how AI agents, powered by CrewAI, automate complex tasks with minimal human input by integrating adaptive workflows, real-time data analysis, and iterative improvements.

Learn how autonomous AI agents automate tasks with minimal supervision, their architecture, applications, risks, and how to build a HackerNews AI news reporter.

This article explores how AI agents are transforming retail by automating customer interactions, optimizing decision-making, and enhancing product recommendations using LLM-driven vector search.

This article explores how agentic RAG enhances retrieval-augmented generation by using AI agents to dynamically refine search strategies, coordinate multiple data sources, and improve response accuracy.

Experimenting with Anthropic's new flagship LLM, Claude 3.7 Sonnet!

Gemini 2.5 Pro Experimental is Google's most advanced AI model to date, featuring multimodal input support, a massive 1 million-token context window, and the ability to solve complex problems.

A practical walkthrough for building OpenAI Agents that use the Model Context Protocol (MCP) to access tools, files, and trace data via Weave.

A practical deep dive into quantization-aware training, covering how it works, why it matters, and how to implement it end-to-end.

Deploy Llama 4 locally or via API with Python scripts. We test multimodal performance against GPT-4o on ChartQA and show how to debug and compare results using Weave.

Retrieval Augmented Thinking (RAT) separates AI reasoning from response generation, improving efficiency, interpretability, and customization by using one model for structured thought and another for the final output.

A hands-on guide to building and comparing memory-based and model-based collaborative filtering systems to quickly evaluate recommendation strategies.

Learn how to build a Reddit sentiment analysis pipeline that uses GPT-4o-mini to extract opinions from real discussions across subreddits—filtering, summarizing, and classifying posts and comments at scale.

The Agent2Agent (A2A) protocol is an open standard that enables autonomous AI agents to securely discover, communicate, and collaborate across platforms. Learn how it works, its core components, and how to implement it.

A hands-on introduction to reinforcement learning, explaining how agents learn from interaction to solve complex decision-making problems - plus practical implementations of Deep Q-learning and Actor-Critic methods in Python.

A hands-on guide to building and evaluating real-time, tool-using AI agents with Azure AI Foundry Agent Service, SerpAPI, and W&B Weave.

Getting set up and running Anthropic's new Claude 4 Sonnet and Opus on your machine in Python using the API.

This article provides a practical introduction to multi-agent reinforcement learning (MARL), explaining its theoretical foundations, key algorithms, and frameworks, and showcasing a custom-coded multi-agent Pong environment with self-play DQN agents to illustrate the opportunities and challenges of training AI agents that interact, compete, or cooperate within shared environments.

Putting Claude 4 Sonnet and Opus to the test

An exploration of whether vector retrieval still makes sense now that LLMs can handle massive context windows.

Google’s Agent Development Kit (ADK) is a modern, modular Python framework for building, orchestrating, and tracing sophisticated AI-powered agents and workflows - supporting diverse models and tools and designed for easy integration, extensibility, and production observability.

Google’s GenAI SDK provides developers with a unified, flexible toolkit to seamlessly integrate advanced generative AI capabilities—including text, image, and video processing—into their applications using the latest Gemini models.

Explore how reinforcement learning with verifiable rewards (RLVR) and GRPO shape LLM reasoning gains, where true improvements lie, plus a practical GRPO training guide.

Explore how reinforcement learning helps AI learn from trial and error, with key algorithms, methods like RLHF, and real-world applications.

This article introduces ART, an open-source framework that simplifies applying reinforcement learning to large language models, enabling developers to efficiently train and deploy LLM agents that continuously improve through experience, as demonstrated with practical code examples including a tic-tac-toe game.

This article explains how reinforcement learning from human feedback (RLHF) is used to train language models that better reflect human preferences, including practical steps, code examples, and evaluation techniques.

AI agents are revolutionizing software automation by leveraging protocols like MCP to directly access databases and developer tools, but this new power comes with significant security risks, as unchecked user input combined with elevated permissions can lead to catastrophic data breaches.

A practical guide to running production AI agents on AWS AgentCore and wiring them to Weave for step-by-step observability, with examples showing setup, instrumentation, and trace-driven debugging.

This article provides a practical, end-to-end walkthrough of building, testing, and evaluating an AI insurance agent using the Agent Development Kit (ADK) and W&B Weave.

These tutorials cover how to evaluate GPT-5’s image generation, coding evals, and automated debugging using W&B Weave.