3 papers
cs.CL2026
The Geometry of Low-Resource Language Representations
Francois Meyer, Jan Buys
The performance gap between low- and high-resource languages in LLMs is widely known, but it remains unclear which internal model factors drive these disparities. In this paper, we…
cs.CL2026
MzansiText and MzansiLM: An Open Corpus and Decoder-Only Language Model for South African Languages
Anri Lombard, Simbarashe Mawere, Temi Aina +5
Decoder-only language models can be adapted to diverse tasks through instruction finetuning, but the extent to which this generalizes at small scale for low-resource languages rema…
cs.CL2025
The Learning Dynamics of Subword Segmentation for Morphologically Diverse Languages
Francois Meyer, Jan Buys
Subword segmentation is typically applied in preprocessing and stays fixed during training. Alternatively, it can be learned during training to optimise the training objective. In…