data contamination
/ˈdeɪtə kənˌtæmɪˈneɪʃən/
test material leaking into training data, inflating benchmark scores
data contamination in a sentence
“The model aced the benchmark through data contamination — it had memorized the answers.”
Origin of data contamination
Latin contaminare to defile by contact
Related Words
prompt leakage
a model revealing its hidden system prompt when coaxed
hallucination
generating plausible-sounding but factually incorrect or fabricated information
sycophancy
over-agreeing with users and telling them what they want to hear rather than the truth
confabulation
filling gaps in knowledge with plausible but invented details
instruction drift
gradually deviating from initial instructions over long conversations
mode collapse
converging to repetitive or generic outputs regardless of varied inputs