paper-with-me

홈 › Papers

A partition cover approach to tokenization

2025-01-08 · Jia Peng Lim, Shawn Tan, Davin Choo, Hady W. Lauw

Tokenization is the process of encoding strings into tokens of a fixed vocabulary size, and is widely utilized in Natural Language Processing applications. The leading tokenization algorithm today is Byte-Pair Encoding (BPE), which formulates the tokenization problem as a compression problem and tackles it by performing sequences of merges. In this work, we formulate tokenization as an optimization objective, show that it is NP-hard via a simple reduction from vertex cover, and propose a polynomial-time greedy algorithm GreedTok. Our formulation naturally relaxes to the well-studied weighted maximum coverage problem which has a simple $(1 - 1/e)$-approximation algorithm GreedWMC. Through empirical evaluations on real-world corpora, we show that GreedTok outperforms BPE and Unigram on compression and achieves a covering score comparable to GreedWMC. Finally, our extensive pre-training for two transformer-based language models with 1 billion parameters, comparing the choices of BPE and GreedTok as the tokenizer, shows that GreedTok achieves a lower bit per byte even when we control for either the total dataset proportion or total training tokens.

📄 PDF Abstract BibTeX arXiv:2501.06246

Code (1)

preferredai/pcatt 공식 구현 jax

Methods 이 논문이 사용한 방법론

BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…

Similar Papers 제목 키워드 기반

A Spitting Image: Modular Superpixel Tokenization in Vision Transformers

2024-08-14 · Marius Aasan, Odd Kolbjørnsen, Anne Schistad Solberg, Adín Ramirez Rivera

Vision Transformer (ViT) architectures traditionally employ a grid-based approach to tokenization independent of the semantic content of an image. We propose a modular superpixel tokenization strategy which decouples tok…

Superpixel Tokenization for Vision Transformers: Preserving Semantic Integrity in Visual Tokens

2024-12-06 · Jaihyun Lew, Soohyuk Jang, Jaehoon Lee, Seungryong Yoo 외

Transformers, a groundbreaking architecture proposed for Natural Language Processing (NLP), have also achieved remarkable success in Computer Vision. A cornerstone of their success lies in the attention mechanism, which …

Superpixels

A Triadic Suffix Tokenization Scheme for Numerical Reasoning

2026-04-13 · Olga Chetverina arxiv

Standard subword tokenization methods fragment numbers inconsistently, causing large language models (LLMs) to lose positional and decimal structure - a primary driver of errors in arithmetic and scientific reasoning. We…

BERT is Not an Interlingua and the Bias of Tokenization

2019-11-01 · WS 2019 11 · Jasdeep Singh, Bryan McCann, Richard Socher, Caiming Xiong

Multilingual transfer learning can benefit both high- and low-resource languages, but the source of these improvements is not well understood. Cananical Correlation Analysis (CCA) of the internal representations of a pre…

ClusteringTransfer Learning

TrajTok: Adaptive Spatial Tokenization for Trajectory Representation Learning

2026-05-19 · Zhen Xiong, Shang-Ling Hsu, Cyrus Shahabi arxiv

Learning generalizable trajectory representations from raw GPS traces remains difficult because the data is continuous, noisy, and irregularly sampled. Spatial tokenization is also challenging: fine grids yield sparse ce…

Representation Learning