#sampling

11 sampling interview questions, asked at Google, Perplexity, Amazon and 8 more companies. Last asked Sep 2026.

+1
Greedy, top-k and top-p decoding for a language modelCodingYou get a toy language model with a small vocabulary whose forward pass returns next-token logits for the tokens so far. Using NumPy, write the generation loop and three ways to…
Last asked Sep 2026
Importance sampling in LLM trainingML designExplain importance sampling: what it is and why it works. Then explain where it is used when training large language models and why it is used there. Expect to go beyond the…
Last asked Aug 2026
Top-p (nucleus) sampling in NumPyCodingImplement top-p sampling in NumPy: from a vector of logits, keep the smallest set of most likely tokens whose probabilities add up to at least p, renormalize them, and sample one…
Last asked Jun 2026
+1
Weighted data batcher with deterministic resumeCodingBuild a sampler that draws weighted batches from a dataset registry and can resume deterministically from a saved offset, including the edge case where the batch size does not…
Last asked May 2026
Substring search and weighted random samplingCodingTwo problems for a machine learning engineer screen: 1. Find where a pattern occurs in a text, and discuss doing better than checking every position. 2. Pick an item at random…
Last asked Mar 2026
Simulate particles from a point source hitting a screenCodingA point source sits some distance in front of an infinitely tall vertical screen and fires particles toward it. Each particle leaves at a random angle drawn uniformly between -π/2…
Last asked Feb 2026
Next-word predictor from bigram countsCodingYou get training sentences as lists of words, for example "a", "b", "c", "a", "s", "d". 1. Count, for every word, which words follow it and how often. 2. predict(word) returns the…
Last asked Feb 2026
Implement multi-head attention and LLM samplingCodingImplement multi-head self-attention from scratch, then the sampling step of a language model: given logits, pick the next token with temperature, top-k and top-p (nucleus)…
Last asked Feb 2026
Speed up a byte tokenizer and estimate token countsCodingYou are given a slow but correct reference tokenizer over bytes. Explain precisely what it does, then write a faster version that returns identical output. Finally, estimate the…
Last asked Jan 2026
Reservoir sampling and a check that it is uniformCodingPick k items uniformly at random from a stream of unknown length in one pass and O(k) memory. Then write a simulation that tests whether your picks are close to uniform, and…
Last asked Nov 2025
+1
Speculative decoding for faster generationCodingImplement speculative decoding: a small draft model proposes tokens and the large model verifies them, accepting matching runs to save forward passes. Reported on 1point3acres'…