Training cutoff
The approximate date after which a model's main training data stops, stated in its documentation. Material created after the cutoff is a ready source of clean items the model is very unlikely to have memorized, making it a cheap contamination probe, though the date is approximate and a later fine-tuning stage can still introduce newer material.
Defined in 2 GAGE programs, which carry 3 distinct definitions of it. The wording above is taught in AI Governance: Applied Mastery.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
The approximate date after which a model's main training data stops, stated in its documentation. Material created after the cutoff is a ready source of clean items the model is very unlikely to have memorized, making it a cheap contamination probe, though the date is approximate and a later fine-tuning stage can still introduce newer material.
The point in time after which a model has no built-in knowledge of events, because its training data ends there. A primary source of outdated-fact errors.
The fixed date up to which a model's training data extends. The model has no knowledge of events, rulings, releases, or facts after this date and will often fabricate a plausible answer rather than admit the gap when asked about recent matters.
Where it is taught
The exact lessons this term appears in. The first 7 topics of every program are free with a free account.
- Evaluating AI Outputs for Accuracy · Practical AI Workflow Design and Prompt Engineering, AI Literacy & Professional Conduct
- What a model cannot know: hallucination produced on demand, then caught · Build Before You Govern, AI Governance: Applied Mastery
- Reproducing a claim: testing a vendor benchmark yourself in an afternoon · Staying Current: The Frontier Discipline, AI Governance: Applied Mastery
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.