arXiv stat.ML
· Papers
Perturbation is All You Need for Extrapolating Language Models
arXiv:2605.04344v2 Announce Type: replace Abstract: This paper develops a statistical theory of extrapolation for large language models, by reinterpreting them through pre-post-additive noise models. In contrast to the standard autoregressive next-token prediction based on an exact prefix, we introduce a perturbation-b