Question bank

4,310 interview questions, answered.

Reference answers, what the interviewer is really testing, how it is graded, and the follow-ups that come next.

Easy 190Medium 2,498Hard 982
FiltersDifficulty, topic, company

4,310 questions

  1. 4105Post-cutoff knowledge with reliabilityAI ML
  2. 4106PPO vs DPO trade-offsAI ML
  3. 4107Project deep-dive + cross-functional disagreementAI ML
  4. 4108RAG, embeddings, reranking walk-throughMediumAI ML
  5. 4109Real-time abuse / rate-limit detectionAI ML
  6. 4110Real-time fraud detectionAI ML
  7. 4111Research presentationAI ML
  8. 4112Research Scientist loop: ML system design + applied researchAI ML
  9. 4113RLHF / DPO post-trainingMediumAI ML
  10. 4114Safety / alignment presence in the loopAI ML
  11. 4115Safety, misuse & content-harm probeAI ML
  12. 4116Scaling-laws critique (Chinchilla re-approach)MediumAI ML
  13. 4117Societal impact and xAI's missionAI ML
  14. 4118Sorted-list intersectionAI ML
  15. 4119Streaming dedup without storing the streamAI ML
  16. 4120System design triple: inference API, GPU scheduling, streamingAI ML
  17. 4121The research-paper take-homeAI ML
  18. 4122The "XAI theory" round (explainability)AI ML
  19. 4123Transformer from scratchMediumAI ML
  20. 4124URL shortenerAI ML
  21. 4125What the patterns mean for your prepAI ML
  22. 4126Why OpenAI over the other labsMediumAI ML
  23. 4127Game LeaderboardHardCodingData
  24. 4128Imagine you're conducting a user survey. What do you do if you can't get 'enough' respondents?MediumTechnical