pith. sign in

You should evaluate your language model on marginal likelihood over tokenisations

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

LangMAP: A Language-Adaptive Approach to Tokenization

cs.CL · 2026-06-22 · unverdicted · novelty 7.0

LangMAP adapts UnigramLM for multilingual use to deliver language-specific tokenization from a shared vocabulary, boosting boundary alignment metrics across natural and programming languages with mixed downstream fine-tuning gains.

citing papers explorer

Showing 1 of 1 citing paper.

  • LangMAP: A Language-Adaptive Approach to Tokenization cs.CL · 2026-06-22 · unverdicted · none · ref 56

    LangMAP adapts UnigramLM for multilingual use to deliver language-specific tokenization from a shared vocabulary, boosting boundary alignment metrics across natural and programming languages with mixed downstream fine-tuning gains.