MultiHashFormer enables hash-based autoregression in LMs by encoding tokens as multi-hash signatures, outperforming standard Transformers at 100M-3B scales while keeping parameter count constant for multilingual expansion.
How can we effectively expand the vocabulary of LLMs with 0.01 GB of target language text? Computational Linguistics, pp.\ 1--40, 11 2025 b
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.CL 3representative citing papers
Low-resource languages are structurally more different from English in LLMs than high- or mid-resource ones, and language-specific post-training alters structures while preserving inter-language relationships.
SSU mitigates catastrophic forgetting in low-resource LLM target-language adaptation by scoring and column-wise freezing source-critical parameters, reducing source degradation to ~3% versus ~20% for full fine-tuning while matching target performance.
citing papers explorer
-
MultiHashFormer: Hash-based Generative Language Models
MultiHashFormer enables hash-based autoregression in LMs by encoding tokens as multi-hash signatures, outperforming standard Transformers at 100M-3B scales while keeping parameter count constant for multilingual expansion.
-
Multilinguality of Large Language Models From a Structural Perspective
Low-resource languages are structurally more different from English in LLMs than high- or mid-resource ones, and language-specific post-training alters structures while preserving inter-language relationships.