https://sankethbk.github.io/blog/tags/sgns/Recent content in Sgns on Sanketh's BlogHugo -- 0.169.0en-usTue, 27 Oct 2026 00:00:00 +0530https://sankethbk.github.io/blog/posts/ml/labs/2026-10-27-lab-2-1-how-word-vectors-learn-to-mean/Tue, 27 Oct 2026 00:00:00 +0530https://sankethbk.github.io/blog/posts/ml/labs/2026-10-27-lab-2-1-how-word-vectors-learn-to-mean/<h1 id="lab-21-how-word-vectors-learn-to-mean--embeddings-from-zero-and-what-they-secretly-compute">Lab 2.1: How Word Vectors Learn to Mean — Embeddings From Zero, and What They Secretly Compute</h1> <p><strong>Language-model internals series.</strong> In the <a href="https://sankethbk.github.io/blog/posts/ml/labs/2026-10-19-lab-1-mnist-from-pixels-to-neural-network/">MNIST series</a> we took apart what weights represent for vision. New series, same ritual — build it from scratch, check gradients numerically, claim nothing without a baseline or a measured number.</p> <p>The question for this lab, in the plainest possible form:</p> <blockquote> <p>A word becomes a row of numbers. How can those numbers possibly <em>mean</em> anything?</p> </blockquote>