Min-k Sampling dynamically truncates the token distribution at each step by detecting semantic cliffs via position-weighted relative logit decay, achieving strict temperature invariance.
InThe Twelfth In- ternational Conference on Learning Representations (ICLR 2024)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Min-$k$ Sampling: Decoupling Truncation from Temperature Scaling via Relative Logit Dynamics
Min-k Sampling dynamically truncates the token distribution at each step by detecting semantic cliffs via position-weighted relative logit decay, achieving strict temperature invariance.