Chinchilla Scaling Laws Quiz

Answers and explanations appear as you take the quiz.

1. What does Chinchilla's finding that parameters and tokens should scale in similar proportions mean?
   • Both should grow with comparable scaling exponents under the studied training-compute objective
   • The numerical token count must equal the parameter count
   • Training data does not affect the fitted loss
   • Every deployment must choose the largest possible model
2. What token-per-parameter ratio does the 70B-parameter, 1.4T-token Chinchilla experiment illustrate?
   • Exactly one token per parameter
   • About two thousand tokens per parameter
   • About twenty training tokens per parameter
   • A ratio determined only by the number of GPUs
3. Which model did Chinchilla outperform despite being much smaller?
   • GPT-2
   • Gopher
   • BERT
   • Mamba
4. What did Chinchilla imply about models like GPT-3 at the time?
   • They were overtrained
   • They had too little compute available
   • They used the wrong tokenizer
   • They were undertrained relative to their size
5. Roughly how was compute approximated in the Chinchilla analysis?
   • Proportional to the number of GPUs only
   • Independent of data size
   • Proportional to parameters times training tokens
   • Proportional to vocabulary size
6. Why might a deployment objective favor a smaller model trained on extra data rather than the training-compute optimum alone?
   • Repeated inference costs can matter alongside the initial training budget
   • Inference has no resource cost for any model
   • The Chinchilla study proved every scaling relationship false
   • More data always reduces both training and serving work to zero
7. Under the approximation that training compute is proportional to N times D, how must D change if N doubles while that compute budget stays fixed?
   • D must also double
   • D must quadruple
   • D must remain unchanged
   • D must approximately halve
8. Which three types of term appear in Chinchilla’s parametric loss fit?
   • Only GPU count, electricity price, and elapsed seconds
   • An irreducible base term, a model-size contribution, and a data-size contribution
   • Only vocabulary size, file size, and model name
   • Only a linear token count with no model-size term