No article text was recorded for this version.
subword-based embedding initialization
Version 1 of 148 papers60 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Fine-tuning transformers: Vocabulary transfer
- Blog: IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effective Domain-Specific Vocabulary Initialization
- Blog: OFA: A Framework of Initializing Unseen Subword Embeddings for Efficient Large-scale Multilingual Continued Pretraining
- Blog: AVocaDo: Strategy for Adapting Vocabulary to Downstream Domain
- Blog: Optimizing Segmentation Granularity for Neural Machine Translation
- Blog: Accelerating Multilingual Language Model for Excessively Tokenized Languages
- Blog: Task-Adaptive Tokenization: Enhancing Long-Form Text Generation Efficacy in Mental Health and Beyond
- Blog: COIN – an Inexpensive and Strong Baseline for Predicting Out of Vocabulary Word Embeddings
- Blog: Exploring Design Choices for Building Language-Specific LLMs
- Blog: From English-Centric to Effective Bilingual: LLMs with Custom Tokenizers for Underrepresented Languages
- Blog: Efficiently Upgrading Multilingual Machine Translation Models to Support More Languages
- Blog: Efficient Continual Pre-training of LLMs for Low-resource Languages
- Blog: AVocaDo: Strategy for Adapting Vocabulary to Downstream Domain
- Blog: Efficient Vocabulary Reduction for Small Language Models
- Blog: TMU NMT System with Japanese BART for the Patent task of WAT 2021
- Blog: Total Recall: a Customized Continual Learning Method for Neural Semantic Parsers
- Blog: ALLaM: Large Language Models for Arabic and English
- Blog: A polyphone BERT for Polyphone Disambiguation in Mandarin Chinese
- Blog: Adapt-and-Distill: Developing Small, Fast and Effective Pretrained Language Models for Domains
- Blog: Enhancing Multilingual ASR for Unseen Languages via Language Embedding Modeling
- Blog: Dynamic data sampler for cross-language transfer learning in large language models
- Blog: SambaLingo: Teaching Large Language Models New Languages
- Blog: WECHSEL: Effective initialization of subword embeddings for cross-lingual transfer of monolingual language models
- Blog: R-BPE: Improving BPE-Tokenizers with Token Reuse
- Blog: NusaBERT: Teaching IndoBERT to be Multilingual and Multicultural
- Blog: Targeted Multilingual Adaptation for Low-resource Language Families
- Blog: End-to-End Autoregressive Retrieval via Bootstrapping for Smart Reply Systems
- Blog: Small Languages, Big Models: A Study of Continual Training on Languages of Norway
- Blog: Enhancing Long-form Text Generation Efficacy with Task-adaptive Tokenization
- Blog: Tuning Multilingual Transformers for Language-Specific Named Entity Recognition
- Blog: Getting the most out of your tokenizer for pre-training and domain adaptation
- Blog: AdaptiVocab: Enhancing LLM Efficiency in Focused Domains through Lightweight Vocabulary Adaptation
- Blog: Evaluating Tokenizer Adaptation Methods for Large Language Models on Low-Resource Programming Languages
- Blog: Fast Vocabulary Transfer for Language Model Compression
- Blog: Entropy-guided Vocabulary Augmentation of Multilingual Language Models for Low-resource Tasks
- Blog: ALLaM: Large Language Models for Arabic and English
- Blog: Enhancing Out-of-Vocabulary Estimation with Subword Attention
- Blog: Exploring Design Choices for Building Language-Specific LLMs
- Blog: UniBridge: A Unified Approach to Cross-Lingual Transfer Learning for Low-Resource Languages
- Blog: Large Vocabulary Size Improves Large Language Models
- Blog: InternSVG: Towards Unified SVG Tasks with Multimodal Large Language Models
- Blog: Adapters for Altering LLM Vocabularies: What Languages Benefit the Most?
- Blog: MIGRATE: Cross-Lingual Adaptation of Domain-Specific LLMs through Code-Switching and Embedding Transfer
- Blog: Second Language (Arabic) Acquisition of LLMs via Progressive Vocabulary Expansion
- Blog: IIIDYT at SemEval-2018 Task 3: Irony detection in English tweets
- Blog: From Tower to Spire: Adding the Speech Modality to a Translation-Specialist LLM
- Blog: Efficient Domain Adaptation of Language Models via Adaptive Tokenization
- Blog: Efficient Low-Resource Language Models Using Tokenizer Transfer
History & citation
Current · Permanent link
- This direction split from Vocabulary Expansion via Subword Embedding Composition (73% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_subword_based_embedding_initialization_5830b028,
title = {subword-based embedding initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/subword-based-embedding-initialization/rdr_5830b02894824916a832ca788670e38a},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}This exact version (version 1, frozen text):
@misc{lacuna_subword_based_embedding_initialization_5830b028_v1,
title = {subword-based embedding initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/subword-based-embedding-initialization/rdr_5830b02894824916a832ca788670e38a/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}