Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: Machine Learning
Jev: How a Non-Generative AI Model Turns Text Into Decisions
Learn how Jev's numeric "System One" AI model works, why it's faster and cheaper than LLMs, and where it fits alongside generative AI in production systems.
7 min readHow Multimodal Models Relate Text, Images and Audio in One Request
Understand what a modality is, how multimodal models map text, images and audio into a shared representation, where that helps, and where visual answers still go wrong.
5 min readRecursive Self-Improvement in AI Agents: What the Experiments Actually Show
A careful look at Gödel Machines, STOP, Gödel Agent, SICA, the Darwin Gödel Machine, SEAL and Hyperagents, and how to judge whether a self-modifying agent truly improves.
9 min readTraining Your Own LLM Judge: From PandaLM and JudgeLM to Prometheus
How finetuned evaluator models are built, from data and rubrics to bias control and meta-evaluation, and a practical recipe for training a domain-specific LLM judge.
39 min readLLM Scaling Laws, GPT-4o and Reinforcement Learning: What the Curves Predict
How to read LLM scaling laws, what Kaplan and Chinchilla found, how labs used them from GPT-3 to GPT-4o, and why reinforcement learning adds a new compute axis.
30 min readFrom GPT-1 to Reasoning Models: What Each Generation Changed for Developers
Trace the GPT family from 2018 pre-training to reasoning models, see which ideas each generation added, and call GPT-4o vision and Structured Outputs from Python.
10 min readBPE Tokenizer From Scratch: Shorter Sequences for a Small GPT Context Window
Write a Byte Pair Encoding tokenizer in Python, train it on WikiText-2, save merges as JSON, and see how subword tokenization stretches a GPT context window.
26 min readRecurrent Depth Explained: Looped Transformers and Hidden Reasoning
How recurrent-depth (looped) transformers reason in hidden state, why labs want them, what eight years of papers show, and what it means for local AI and oversight.
12 min readDebugging a Small GPT in PyTorch: Tests That Isolate Each Failure
A stage-by-stage workflow for debugging a character-level GPT in PyTorch, from token IDs and target shifting to gradients, NaN losses and checkpoints.
30 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.