Towards Data Science

2 min read Original article ↗

  • what it costs, what it gains and the three mistakes that I make

  • I got tired of copying files into an AI chat just to get feedback. So…

Latest

  • How a simple choice shapes exploration, safety, and efficiency

  • Using DSPy to automatically create, evaluate, and optimize your prompts

  • Python tutorial for fine-tuning a Mistral Small 3.1 on an imbalanced training set to classify…

  • Abacus.AI and the case for unified AI workflows

  • In Part 1 of this series, we introduced Chronos-2, a time-series foundation model. We got…

  • Image by author.

    When images, mosaics, and data cubes exist in abundance, but field labels are expensive, rare,…

  • Understanding how FPN allows deep learning models detecting small objects and how to implement it…

  • A look at the real-world value of online graduate AI programs, combining hard data with…

  • Testing fourteen engines on ninety-three human documents

Editor’s Picks

  • AI does not decide who gets fired. Companies do.

  • A comprehensive guide to optimizing LLM inference by eliminating padding overhead with hardware-aware sequence packing.

  • How to set the rules that keep agents effective and out of trouble

  • The barriers to building have collapsed. That shifts the bottleneck to ownership, validation, taste, and…

  • The data multiverse

    Research projects in the age of AI

  • How Knives Out teaches Bayesian thinking (without you realizing it) 

  • As AI gets smarter, the real differentiator may be how well humans regulate their own…

  • Part 1: A practitioner’s walkthrough of univariate, multivariate, covariate-informed, and cold-start forecasting.

  • A diffusion-inspired framework for stress-testing and denoising LLM-as-a-Judge pipelines, applied to safety-critical driving video.

The Variable Newsletter

  • Authors can now benefit from updated earning tiers and a higher article cap

  • Sorting through the good, bad, and ambiguous aspects of vibe coding

Deep Dives

  • Why this image: an instrument for examining the small print, what this article does to the casual claim that RAG is an ML problem, looking closely at where the ML toolkit fits and where it does not.

    Enterprise Document Intelligence [Vol.1 #3] – Why the ML toolkit (hyperparameter sweeps, train/test splits, explainability…

  • Enterprise Document Intelligence [Vol. 1 #2] Why the same vector search that handles synonyms and…

  • Most engineers see quantization as shrinking vectors. TurboQuant asks a harder question: can you shrink…

  • Enterprise Document Intelligence [Vol. 1 #1] The smallest version of RAG that actually works, on…

  • Most RAG systems are optimized for answer quality, not cost—and that blind spot gets expensive…

  • A step-by-step journey from calculus-based optimization to Stochastic Gradient Descent