—
Diagnose model evaluation failures
Tutor me on https://mlprep.iwase.dev/item/mit-6-036-s2019-midterm-q4/. This is a practice item. If window.mlPrepAgent is available, read attempts for item mit-6-036-s2019-midterm-q4 before tutoring. Inspect my attempt, keep the item ID, and do not reveal the full answer first. After a real attempt, append its record with recorded_by agent, include agent_session_id when available, and read it back.Uses observed train and held-out behavior to separate overfitting, insufficient expressivity, data scarcity, and distribution shift.
Previous attempts
Loading attempts saved in this browser…