paper-with-me

홈 › Papers

Modeling the Unigram Distribution

2021-06-04 · Findings (ACL) 2021 8 · Irene Nikkarinen, Tiago Pimentel, Damián E. Blasi, Ryan Cotterell

The unigram distribution is the non-contextual probability of finding a specific word form in a corpus. While of central importance to the study of language, it is commonly approximated by each word's sample frequency in the corpus. This approach, being highly dependent on sample size, assigns zero probability to any out-of-vocabulary (oov) word form. As a result, it produces negatively biased probabilities for any oov word form, while positively biased probabilities to in-corpus words. In this work, we argue in favor of properly modeling the unigram distribution -- claiming it should be a central task in natural language processing. With this in mind, we present a novel model for estimating it in a language (a neuralization of Goldwater et al.'s (2011) model) and show it produces much better estimates across a diverse set of 7 languages than the na\"ive use of neural character-level language models.

📄 PDF Abstract BibTeX arXiv:2106.02289

Code (1)

irenenikk/modelling-unigram 공식 구현 pytorch

Tasks

Form

Similar Papers 제목 키워드 기반

Conditional Unigram Tokenization with Parallel Data

2025-07-10 · Gianluca Vico, Jindřinch Libovický arxiv

We introduce conditional unigram tokenization, a novel approach that extends unigram tokenization by conditioning target token probabilities on source-language tokens from parallel data. Given a fixed source tokenizer, o…

Machine Translation

Semantic Composition and Decomposition: From Recognition to Generation

2014-05-30 · Peter D. Turney

Semantic composition is the task of understanding the meaning of text by composing the meanings of the individual words in the text. Semantic decomposition is the task of understanding the meaning of an individual word b…

Semantic Composition

Where to cut, how deep: BPE and Unigram-LM on chemistry SMILES

2026-07-06 · Hunter Heidenreich hf

Every chemical language model reading SMILES begins with a tokenizer, yet the field has inherited byte-pair encoding (BPE) from natural language with little scrutiny. In natural language, BPE's principal alternative, Uni…

Byte Pair Encoding is Suboptimal for Language Model Pretraining

2020-04-07 · Findings of the Association for Computational Linguistics 2020 · Kaj Bostrom, Greg Durrett

The success of pretrained transformer language models (LMs) in natural language processing has led to a wide range of pretraining setups. In particular, these models employ a variety of subword tokenization methods, most…

Language ModelingLanguage Modelling

A Natural Bias for Language Generation Models

2022-12-19 · Clara Meister, Wojciech Stokowiec, Tiago Pimentel, Lei Yu 외

After just a few hundred training updates, a standard probabilistic model for language generation has likely not yet learnt many semantic or syntactic rules of natural language, making it difficult to estimate the probab…

Machine TranslationText Generation