CIST uses per-sample adaptive temperatures for both teacher and student in knowledge distillation to ensure consistent entropy in soft labels and reports gains on vision and language tasks.
Promptkd: Distilling student-friendly knowledge for generative language models via prompt tuning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Consistently Informative Soft-Label Temperature for Knowledge Distillation
CIST uses per-sample adaptive temperatures for both teacher and student in knowledge distillation to ensure consistent entropy in soft labels and reports gains on vision and language tasks.