Evaluating Subword Tokenization: Alien Subword Composition and OOV Generalization Challenge
The popular subword tokenizers of current language models, such as Byte-Pair Encoding (BPE), are known not to respect morpheme boundaries, which affects the downstream performance of the models. While many improved tokenization algorithms have been proposed, their evaluation and cross-comparison is still an open problem. As a solution, we propose a combined intrinsic-extrinsic evaluation framework for subword tokenization. Intrinsic evaluation is based on our new UniMorph Labeller tool that classifies subword tokenization as either morphological or alien. Extrinsic evaluation, in turn, is performed via the Out-of-Vocabulary Generalization Challenge 1.0 benchmark, which consists of three newly specified downstream text classification tasks. Our empirical findings show that the accuracy of UniMorph Labeller is 98%, and that, in all language models studied (including ALBERT, BERT, RoBERTa, and DeBERTa), alien tokenization leads to poorer generalizations compared to morphological tokenization for semantic compositionality of word meanings.
Code (1)
Tasks
text-classificationText ClassificationMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Assessing the Importance of Frequency versus Compositionality for Subword-based Tokenization in NMT
Subword tokenization is the de facto standard for tokenization in neural language models and machine translation systems. Three advantages are frequently cited in favor of subwords: shorter encoding of frequent tokens, c…
de-enMachine TranslationNMTTranslationFrom Words to Music: A Study of Subword Tokenization Techniques in Symbolic Music Generation
Subword tokenization has been widely successful in text-based natural language processing (NLP) tasks with Transformer-based models. As Transformer models become increasingly popular in symbolic music-related studies, it…
Music GenerationWords, Subwords, and Morphemes: What Really Matters in the Surprisal-Reading Time Relationship?
An important assumption that comes with using LLMs on psycholinguistic data has gone unverified. LLM-based predictions are based on subword tokenization, not decomposition of words into morphemes. Does that matter? We ca…
Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation
Subword tokenization is an essential part of modern large language models (LLMs), yet its specific contributions to training efficiency and model performance remain poorly understood. In this work, we decouple the effect…
Charformer: Fast Character Transformers via Gradient-based Subword Tokenization
State-of-the-art models in natural language processing rely on separate rigid subword tokenization algorithms, which limit their generalization ability and adaptation to new settings. In this paper, we propose a new mode…
Inductive BiasLinguistic AcceptabilityNatural Language InferenceParaphrase Identification+2