pith. sign in

Transformer-based Causal Language Models Perform Clustering

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.CL 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

Understanding the Prompt Sensitivity

cs.CL · 2026-04-20 · unverdicted · novelty 5.0

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

Showing 1 of 1 citing paper.

  • Understanding the Prompt Sensitivity cs.CL · 2026-04-20 · unverdicted · none · ref 42

    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.