← 💻 Coding & Engineering CraftNEXT IN CODING & ENGINEERING CRAFTFast and Slow Pointers (Floyd's Cycle Detection)→
Core
Numerical Stability in Code
Numerical stability is writing arithmetic so floating-point error and overflow do not corrupt the result, which matters because naive ML math (softmax, cross-entropy, variance) silently returns NaN or wrong gradients. Applied AI interviews probe it because the fixes (log-sum-exp, max-subtraction, working in log-space) are small code changes that separate engineers who have shipped training loops from those who have only called library functions.
a free account unlocks the core curriculum tier · no card
COURSES COVERING THIS TOPIC
No lesson covers this one directly yet. These teach the surrounding topic from the beginning.
RELATED CONCEPTS
PRACTICE THIS IN REAL QUESTIONS
Coding & DSAImplement a numerically stable softmax and cross-entropy loss from scratch.→LLM & GenAI FundamentalsWhy do transformers scale attention scores by 1/√d_k, and what breaks if you skip it?→Coding & DSAImplement logistic regression from scratch in NumPy: forward pass, loss, and gradient descent.→LLM & GenAI FundamentalsWhat is perplexity, and what are its limits as a language-model metric?→Machine Learning & Data ScienceCompare activation functions (sigmoid, tanh, ReLU, GELU, softmax) and when to use each.→Machine Learning & Data ScienceHow do you choose a loss function (MSE, MAE, Huber, cross-entropy, focal, contrastive)?→
COMPANIES THAT ASSUME THIS