Hallucination
Also called hallucinations
Fluent, confident output that is false or not supported by the input.
Interview questions about Hallucination
- What is a hallucination, and why can a next-token predictor produce one so fluently?
It's fluent, confident output that is false or not grounded in any source.
- Why do long generated answers leave more room for hallucination than one-word answers?
Every generated token becomes context for the next, and nothing marks it as the model's own guess, so each step is another chance to go wrong.
- You're shipping a support bot that answers from your company's policy documents. What do you put in place to reduce hallucination, and how do you evaluate it?
Ground answers in retrieved policy text. Context steers strongly: in our toy, one true sentence moved " Canberra" from 0.0436 to 0.8248, though that mostly…
- How would you measure an LLM's answer quality and detect hallucinations against the context it was given? Then implement the evaluation component.
Quality is task-specific: reference answers where they exist, a rubric judge, human spot checks.
- Design a voice assistant whose answers stay grounded. How would you reduce hallucinations?
Voice adds two problems: errors start at speech recognition, and the user can't see sources.
- How would you use LLMs and vision-language models in autonomous driving, and could an LLM actually drive the car given hallucination and real-time latency limits?
Not in the control loop. Control runs many times a second and must be verifiable; an LLM's latency varies, and a hallucinated object or rule has no safe…
- For a medical conversational AI, how would you source and split data, choose between prompting, fine-tuning and RAG, and evaluate for hallucination and unsafe advice?
Use vetted clinical guidelines and de-identified, consented conversations, split by patient so nothing leaks into the test set.
- What is retrieval-augmented generation, and how does it reduce hallucination?
Before the model answers, retrieve relevant passages from your documents and put them in the prompt, so it answers from evidence in its context, not memory…