andresnamm/blogg-math-skills

Use when the user asks about linear algebra or geometry, e.g. vectors, inner/dot product, angle between vectors, coordinate systems, right-handed vs left-handed, translation, rotation, homogeneous transformation matrix, camera coordinates, orthographic or perspective projection, viewport transform, intrinsics, z-depth vs Euclidean depth, LiDAR projection.

Use when the user asks about calculus, e.g. derivative, limit, partial derivative, directional derivative, gradient, steepest ascent/descent, chain rule, rate of change, linear approximation, why a derivative is a sum.

Use when the user asks about neural networks, e.g. backpropagation, micrograd, computation graph, gradient accumulation, chain rule in training, convolution, kernel gradient, CNN layers, batch normalization, YOLO architecture, loss gradients.

Use when the user asks about statistics or probability, e.g. normality, Q-Q plot, Shapiro-Wilk weights or W, p-values, variance, t-test, Wilcoxon, bootstrap, probability space, Bayes, Bayesian learning, normal distribution, density function (PDF), mean/std, classification metrics (accuracy, precision, recall, F1, confusion matrix).