amazon logo
Amazon· Applied Data Scientist

Amazon | Interview (1st Round) | Bengaluru | Applied Data Scientist

Location
Bengaluru
Experience
4 years
Posted
22 Jun 2025
297 reads

Summary

I recently had my first online interview at Amazon’s Bengaluru office for an Applied Data Scientist role, focusing on theoretical understanding of LLMs, transformer internals, and general ML/statistics concepts. A further round focusing on coding and resume projects is scheduled.

How it went

Interview Experience

I recently had my first online interview at Amazon’s Bengaluru office recently in May 2025. I have 4 years of industry experience as a Data Scientist in a mid sized company. The round was focused on theoretical understanding of large language models (LLMs), transformer internals, and general ML/statistics concepts.

Update

Further round scheduled which would focus on the coding aspects alongwith deep dive into resume projects.

Questions asked (4)

  1. 01

    QKV in BERT vs GPT

    Other

    I was asked to explain the concept of QKV (Query, Key, Value) in the attention mechanism, and specifically how it’s implemented and used differently in BERT vs GPT.

    Follow-up: Why does GPT only use causal masking? What architectural changes exist in decoder-only vs encoder-decoder models?

  2. 02

    Point Estimators, Bias-Variance Tradeoff

    Other

    They wanted to check my statistical intuition. I was asked to define a point estimator, and explain the bias-variance tradeoff with relevant examples.

    Follow-up: What would you prefer in a low-data regime: high bias or high variance model?

  3. 03

    Source of Stochasticity in LLMs

    Other

    What are the key sources of stochastic behavior in LLMs during inference?

    Follow-up: How does temperature sampling and top-k/top-p affect generation?

  4. 04

    How do LLMs perform so well with little data?

    Other

    This was more of a conceptual discussion. I was asked:

    “How do LLMs generalize well with very little task-specific data during fine-tuning or prompting?”

    Follow-up: Why doesn’t the classical bias-variance limitation seem to apply here?

Know someone preparing? Send it to them:WhatsAppLinkedInX

Interviewed at Amazon recently?

Write up your rounds while you still remember the questions. It helps the next person more than you would think.

Share your experience
Discussion 0

Questions for the author, or how your own interview compared.

Join the discussion

Sign in with Google to comment or ask the author a question.

No comments yet

Be the first to share your thoughts and start the discussion!