Concern 86 · Working draft · editorial review in progress
Label drift
Label shift usually means a change in how common the outcome categories are, rather than a change in what their names mean.
Understanding the concern
For example, the proportion of spam in an inbox may rise while the characteristics of spam remain similar. A system's expectations about how often each category occurs may then need adjustment.
Changing the definition of spam is a different issue, closer to concept or annotation-policy change. Keeping these meanings separate helps a team choose the right test and response. Corrections for label shift rely on assumptions that need checking.
An example
A sorting service receives many more urgent requests than it did during training, although the meaning of urgent has not changed.
Illustrative unless identified as a documented case.
Words in plain English
- Annotation
- Human work that labels or describes examples for an AI system. Labels can involve judgment and can be wrong.
- Drift
- A change over time in data or real-world conditions that can make a system less reliable.
Evidence and its limits
The paper formally defines label shift as a change in label frequency under a stated conditional-distribution assumption.
Lipton, Wang and Smola · ICML 2018 · Research study
Detecting and Correcting for Label Shift with Black Box Predictors ↗ (opens in a new tab)Defines changes in outcome frequencies under a conditional-distribution assumption, rather than changes in label meanings.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Did the category frequencies change, or did the rules for assigning the categories change?
Keep track of different views, what needs checking, and any next step in your own discussion notes.
Open discussion notes →Principles for the conversation
Before you finish · reflect together
Virtues and principles to consider
Virtues are qualities we practice in how we treat people and make choices. They are not scores for judging someone. Use these ideas to explore what a thoughtful response could look like.
Humility
Humility means recognizing the limits of our knowledge and being willing to learn from others. Expertise can be valuable without answering every question.
State where evidence is weak, invite criticism, and listen to people with different experience. Leave room to revise a decision when new information appears.
Ask together: What might we be missing, and who could help us see it?
Prudence
Prudence means using careful judgment before and during action. It balances worthwhile opportunities with the need to avoid harm that could have been anticipated.
Start with a limited trial, set stopping points, and prefer choices that can be undone when uncertainty is high. Review what happens rather than assuming the original plan remains sound.
Ask together: What is a careful next step, and when should we stop or change course?
Truthfulness
Truthfulness means making claims that match the evidence and being clear about what we do not know. It includes correcting mistakes rather than protecting a polished story.
Check sources, label guesses, and distinguish a trial result from a proven benefit. If AI helped produce a claim, someone still needs to take responsibility for checking it.
Ask together: What can we honestly say, and what still needs checking?
Editorial ratings
These are editorial judgments from the register, not measured probabilities or predictions about your situation.
About the register and scoring →More ways to explore
Search for more writing
For more specific or recent material, open a search below or ask for a web summary. AI-generated summaries can contain mistakes; open the original articles to check them.