https://nalar.dev/tags/distribution-shift/ Recent content in Distribution Shift on Nalar Hugo en-us Mon, 01 Jan 0001 00:00:00 +0000 https://nalar.dev/compare-model-input-distributions-with-maximum-mean-discrepancy/ Fri, 11 Sep 2026 00:00:00 +0700 https://nalar.dev/compare-model-input-distributions-with-maximum-mean-discrepancy/ <h1 id="compare-model-input-distributions-with-maximum-mean-discrepancy">Compare Model Input Distributions with Maximum Mean Discrepancy</h1> <p>A model can keep returning valid outputs while the data reaching it gradually changes. An image service may receive photos from new cameras. A text classifier may see a new mix of topics. An embedding pipeline may start processing documents from a different source. Accuracy can degrade even though the model binary, API, and serving code are unchanged.</p> <p><strong>Maximum mean discrepancy (MMD)</strong> gives you a way to compare two samples without reducing each one to a few statistics such as a mean and variance. It uses pairwise similarities to produce a single discrepancy score. That makes it useful as one signal for distribution shift, especially when inputs or representations are vectors.</p>