LLM Agents Interview Questions, issue 1, Feb 23, 2026
The Privacy Scaling Trap
Senior AI Engineer interview at NVIDIA, and the interviewer asks:
“Your team just upgraded an internal LLM from a 7B to a 70B parameter model using the exact same training dataset and 100k step schedule. You expect a reasoning bump, but SecOps flags a 400% spike in PII ( Personally Identifiable Information ) extraction via simple prompting. Why does scaling up independently degrade privacy, and how do you fix it without rolling back?”
Don’t say: “The larger model overfit because we didn't increase the dataset size. We need to apply early stopping, increase dropout, or just run regex data scrubbing on the corpus.”
Increasing capacity without increasing data diversity shifts optimization toward memorization modes that validation loss won’t detect but adversarial prompting will expose.
The full answer, with the mechanism and the arithmetic, is for paid subscribers on Substack.