https://nalar.dev/tags/loss-functions/ Recent content in Loss Functions on Nalar Hugo en-us Sun, 06 Sep 2026 00:00:00 +0700 https://nalar.dev/understand-cross-entropy-loss-for-classification/ Sun, 06 Sep 2026 00:00:00 +0700 https://nalar.dev/understand-cross-entropy-loss-for-classification/ <p>A classifier needs more than a way to count correct answers. During training, it needs a signal that says not only whether a prediction was wrong, but also how the model&rsquo;s scores should change.</p> <p>Suppose the correct class is <code>cat</code>. A model that assigns <code>cat</code> probability 0.49 and another class 0.51 is wrong, but it is close to the decision boundary. A model that assigns <code>cat</code> probability 0.001 is also wrong, and much more confident in that mistake. Treating those predictions as equally bad throws away useful information.</p>