Paper proposes graduated disclosure for mixed-sensitivity LLM answers
A reward-based generator cuts unauthorized leakage, including paraphrases, while keeping utility on enterprise and clinical tests.
Stories per day · 30 days
Relatedaccess-control1benchmark1decision-models1enterprise1graduated-disclosure1privacy1quantization1
2 stories, newest first · one card per story
Stories tagged #information-leakage
TUPLE-Bench finds typed decision outputs leak withheld causes
Across nine models, 91.88% to 99.93% of responses revealed a withheld cause through selection or scores.