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. 2185Detecting distribution shift in production LLM trafficHardAI ML
  2. 2186Diagnosing bias vs variance by experimentMediumAI ML
  3. 2187Document chunking strategies for RAGMediumAI ML
  4. 2188DPO vs RLHF/PPO: when would you pick each?HardAI ML
  5. 2189Evaluating reasoning models vs chat modelsHardAI ML
  6. 2190Fine-tuning vs RAG vs prompt engineeringHardAI ML
  7. 2191FlashAttention and why it mattersHardAI ML
  8. 2192Grid search vs random search vs Bayesian optimizationMediumAI ML
  9. 2193Guardrails for agents that call external APIsHardAI ML
  10. 2194Handling severe class imbalance in productionMediumAI ML
  11. 2195How do you evaluate a RAG system?HardAI ML
  12. 2196How do you evaluate an agent?HardAI ML
  13. 2197Knowledge distillation for LLMsHardAI ML
  14. 2198L1 vs L2 regularizationEasyAI ML
  15. 2199Latency vs throughput vs cost: choosing batch sizesHardAI ML
  16. 2200LLM-as-judge: when it works, when it failsHardAI ML
  17. 2201LLM observability vs classical ML observabilityHardAI ML
  18. 2202LoRA vs full fine-tuningMediumAI ML
  19. 2203Measuring hallucination rate without expensive human evalHardAI ML
  20. 2204Mitigating hallucinations in RAGHardAI ML
  21. 2205Model Context Protocol (MCP)MediumAI ML
  22. 2206Monitoring RAG in productionHardAI ML
  23. 2207Multi-agent orchestration: when does it help?HardAI ML
  24. 2208Overfitting: detection and preventionEasyAI ML