CascadeFormer tapers Transformer width with depth based on gradient fan-in asymmetry to match uniform baselines in perplexity while cutting latency 8.6%.
arXiv preprint arXiv:2206.00330 , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
LLMs disperse meaning-preserving prompts internally instead of clustering them, which produces an excessively high upper bound on output log-probability differences via Taylor expansion and Cauchy-Schwarz.
citing papers explorer
-
CascadeFormer: Depth-Tapered Transformers Motivated by Gradient Fan-in Asymmetry
CascadeFormer tapers Transformer width with depth based on gradient fan-in asymmetry to match uniform baselines in perplexity while cutting latency 8.6%.
-
Understanding the Prompt Sensitivity
LLMs disperse meaning-preserving prompts internally instead of clustering them, which produces an excessively high upper bound on output log-probability differences via Taylor expansion and Cauchy-Schwarz.