Subword Segmental BabyLMs: Learning to Tokenise for Sample-Efficient Pretraining

2026-09-01

AI Summary

本文通过开发两种新的子词分段语言模型SubSegGPT和SubSegDeBERTa，在训练过程中学习标记化，以提高BabyLM预训练的样本效率。

Abstract

In the standard LM training pipeline, subword tokenisation is applied as a preprocessing step. Subword segmental language modelling is an alternative paradigm in which tokenisation is learned during training, allowing the model to discover subword units that optimise its training objective. In this paper, we present our submission to the 2026 BabyLM Challenge, for which we develop two new subword segmental LMs: SubSegGPT and SubSegDeBERTa. SubSegGPT is a decoder-only model that learns tokenisation during autoregressive pretraining. SubSegDeBERTa is an encoder-based model that jointly learns to generate and tokenise masked words. We train both for the Strict and Strict-small tracks. Our top submission to Strict is SubSegDeBERTa, which achieves notable gains in zero-shot evaluation. Our top submission to Strict-small is SubSegGPT, which outperforms tokenisation-based baselines. Our results show that learnable subword tokenisation can improve sample-efficiency for BabyLM pretraining. We analyse the subword learning dynamics of our models and find that tokenisation gradually converges on subword units that balance morphological alignment and fine-grained segmentation.

Problem

• subword tokenisation
• sample-efficiency
• pretraining

Innovation

• learnable subword tokenisation
• subword segmental language modelling
• sample-efficiency