Weights & Biases Quiz

Answers and explanations appear as you take the quiz.

1. What core problem in machine learning does Weights & Biases primarily address?
   • Tracking and reproducing experiments and their metrics
   • Compressing neural network weights for mobile devices
   • Generating synthetic training data automatically
   • Encrypting model weights for secure deployment
2. Which two Python calls are the typical entry point for logging to W&B?
   • wandb.train() and wandb.deploy()
   • wandb.init() and wandb.log()
   • model.fit() and model.save()
   • wandb.connect() and wandb.send()
3. What is the purpose of W&B Sweeps?
   • Cleaning corrupted training data
   • Deleting old experiment logs to save storage
   • Automated hyperparameter search across many runs
   • Streaming video of the training GPU
4. What do W&B Artifacts let you do?
   • Render 3D visualizations of neural networks
   • Automatically write research papers
   • Convert models to a different framework
   • Version datasets and models so results can be reproduced
5. Which company acquired Weights & Biases in March 2025?
   • CoreWeave
   • NVIDIA
   • Microsoft
   • Hugging Face
6. What is W&B Weave designed for?
   • Generating embeddings for search
   • Tracing and evaluating LLM applications
   • Managing Kubernetes clusters
   • Labeling images for supervised learning
7. What can a W&B Table help a team compare?
   • Only GPU temperatures from a single machine
   • Only source-code branches with no logged data
   • Predictions or examples across models, epochs, or individual cases
   • Only user accounts and billing plans
8. What does W&B record when a run uses a logged artifact as an input?
   • A guarantee the model will reproduce identical results on all hardware
   • An automatic change to the dataset contents
   • A license grant for every file
   • A versioned input relationship that helps trace run lineage