mlsanityData to 14 Oct 2026

Learning Beyond Limits: Multitask Learning and Synthetic Data for Low-Resource Canonical Morpheme Segmentation

arXiv 2505.16800 ↗First posted 22 May 2025cs.CLcited by 0

On the map: Synthetic data and self-improvement for LLMs → LLM synthetic data generation

Abstract (arXiv)

We introduce a transformer-based morpheme segmentation system that augments a low-resource training signal through multitask learning and LLM-generated synthetic data. Our framework jointly predicts morphological segments and glosses from orthographic input, leveraging shared linguistic representations obtained through a common documentary process to enhance model generalization. To further address data scarcity, we integrate synthetic training data generated by large language models (LLMs) using in-context learning. Experimental results on the SIGMORPHON 2023 dataset show that our approach significantly improves word-level segmentation accuracy and morpheme-level F1-score across multiple low-resource languages.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in LLM synthetic data generation, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 5 of its 14 references that are in the map. First posted on arXiv 22 May 2025.

Builds on

5 of 14 references.

References · 5

  1. 2017Attention Is All You Needcited by 60,361
  2. 2019fairseq: A Fast, Extensible Toolkit for Sequence Modelingcited by 1,488
  3. 2016Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translationcited by 469
  4. 2016Sequence-to-Sequence Learning as Beam-Search Optimizationcited by 127
  5. 2020Applying the Transformer to Character-level Transductioncited by 43

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2020Tackling the Low-resource Challenge for Canonical Segmentation
  2. 2024Boosting the Capabilities of Compact Models in Low-Data Contexts with Large Language Models and Retrieval-Augmented Generation
  3. 2026Hybrid Neural-LLM Pipeline for Morphological Glossing in Endangered Language Documentation: A Case Study of Jungar Tuvan
  4. 2025BabyLM's First Words: Word Segmentation as a Phonological Probing Task
  5. 2024VLMs meet UDA: Boosting Transferability of Open Vocabulary Segmentation with Unsupervised Domain Adaptation
  6. 2022Open-vocabulary Semantic Segmentation with Frozen Vision-Language Models
  7. 2025Segmentation as A Plug-and-Play Capability for Frozen Multimodal LLMs
  8. 2025Evaluating Morphological Alignment of Tokenizers in 70 Languages