SHIELD dataset and distilled DeBERTa v3 model achieve 0.88 micro precision and 0.86 recall on PHI de-identification while matching teacher performance on structured categories.
Targeted Error Correction in Knowledge Distillation : Small Language Models Surpass GPT , November 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SHIELD: A Diverse Clinical Note Dataset and Distilled Small Language Models for Enterprise-Scale De-identification
SHIELD dataset and distilled DeBERTa v3 model achieve 0.88 micro precision and 0.86 recall on PHI de-identification while matching teacher performance on structured categories.