Related papers: A Formal Perspective on Byte-Pair Encoding

Byte Pair Encoding (BPE) tokenizers, widely used in Large Language Models, face challenges in multilingual settings, including penalization of non-Western scripts and the creation of tokens with partial UTF-8 sequences. Pretokenization,…

The success of pretrained transformer language models (LMs) in natural language processing has led to a wide range of pretraining setups. In particular, these models employ a variety of subword tokenization methods, most notably byte-pair…

Tokenization is the first step in modern neural language model pipelines where an input text is converted to a sequence of subword tokens. We introduce from first principles a finite-state transduction framework which can efficiently encode…

Tokenization is a critical preprocessing step in large language model pipelines, yet widely-used implementations remain CPU-bound and suboptimal for batch inference workflows on GPU. We present BlockBPE, a parallel GPU implementation of…

Enumerating maximal $k$-biplexes (MBPs) of a bipartite graph has been used for applications such as fraud detection. Nevertheless, there usually exists an exponential number of MBPs, which brings up two issues when enumerating MBPs, namely…

Binary segmentation is the classic greedy algorithm which recursively splits a sequential data set by optimizing some loss or likelihood function. Binary segmentation is widely used for changepoint detection in data sets measured over space…

Belief Propagation (BP) is a popular, distributed heuristic for performing MAP computations in Graphical Models. BP can be interpreted, from a variational perspective, as minimizing the Bethe Free Energy (BFE). BP can also be used to solve…

This paper introduces Dynamic Programming Encoding (DPE), a new segmentation algorithm for tokenizing sentences into subword units. We view the subword segmentation of output sentences as a latent variable that should be marginalized out…

It was recently shown that the problem of decoding messages transmitted through a noisy channel can be formulated as a belief updating task over a probabilistic network [McEliece]. Moreover, it was observed that iterative application of the…

While subword tokenizers such as BPE and WordPiece are typically used to build vocabularies for NLP models, the method of decoding text into a sequence of tokens from these vocabularies is often left unspecified, or ill-suited to the method…

We analyse the performance of several iterative algorithms for the quantisation of a probability measure $\mu$, based on the minimisation of a Maximum Mean Discrepancy (MMD). Our analysis includes kernel herding, greedy MMD minimisation and…

Subword tokenization methods, such as Byte-Pair Encoding (BPE), significantly impact the performance and efficiency of large language models (LLMs). The standard approach involves training a general-purpose tokenizer that uniformly…

We propose a novel greedy algorithm for the support recovery of a sparse signal from a small number of noisy measurements. In the proposed method, a new support index is identified for each iteration based on bit-wise maximum a posteriori…

Tokenisation is an integral part of the current NLP pipeline. Current tokenisation algorithms such as BPE and Unigram are greedy algorithms -- they make locally optimal decisions without considering the resulting vocabulary as a whole. We…

We explore threshold vocabulary trimming in Byte-Pair Encoding subword tokenization, a postprocessing step that replaces rare subwords with their component subwords. The technique is available in popular tokenization libraries but has not…

The Maximum s-Bundle Problem (MBP) addresses the task of identifying a maximum s-bundle in a given graph. A graph G=(V, E) is called an s-bundle if its vertex connectivity is at least |V|-s, where the vertex connectivity equals the minimum…

The Partitioning Min-Max Weighted Matching (PMMWM) problem is an NP-hard problem that combines the problem of partitioning a group of vertices of a bipartite graph into disjoint subsets with limited size and the classical Min-Max Weighted…

In neural machine translation (NMT), it is has become standard to translate using subword units to allow for an open vocabulary and improve accuracy on infrequent words. Byte-pair encoding (BPE) and its variants are the predominant approach…

Subword tokenization methods like Byte Pair Encoding (BPE) are widely used in large language models due to their balance of vocabulary compactness and representational power. However, they suffer from inefficiencies in representing rare…

Binary Neural Network (BNN) represents convolution weights with 1-bit values, which enhances the efficiency of storage and computation. This paper is motivated by a previously revealed phenomenon that the binary kernels in successful BNNs…