Top 50 Neural Networks Interview Questions
The most frequently asked Neural Networks questions across all roles and companies, ranked by real interview frequency. Updated daily.
Implement a one-hidden-layer neural network with forward propagation and batch gradient descent for binary classification.
Bell
ClickUp
OneMagnifyImplement scaled multi-head attention with stable softmax and causal masking for autoregressive decoding.
Scale
MetaCompute gradients for a two-layer neural network using vectorized backpropagation and softmax cross-entropy.
OpenAIImplement numerically stable scaled dot-product self-attention with optional causal masking using pure language.
AppfolioConstruct a shared-encoder neural network specification whose learned head sums two MNIST digit representations.
Calculate flattened features, trainable parameters, batch updates, and output shapes for a PyTorch training loop.
Torc RoboticsImplement a deterministic binary decision tree that selects Gini-optimal feature thresholds and predicts labels for test samples.
Applied MedicalImplement numerically stable softmax cross-entropy loss and its logits gradient using only tensor operations.
Tempus AISign up to see every question
Create a free account to unlock this list and practice real interview questions.