Speculative decoding enhances the efficiency of large language models (LLMs) by leveraging a draft model to draft for a larger target model to review. However, drafting in speculative decoding involves slow autoregressive generation and generating tokens of different importance with the same time allocation. These two inefficiencies lead to its suboptimal performance. To address this issue, we introduce Cascade Speculative Drafting (CS. Drafting), a novel approach that employs two types of cascades. The Vertical Cascade eliminates autoregressive generation from neural models. The Horizontal Cascade constitutes efficient time allocation in drafting with its optimality supported by our theoretical analysis. Combining both cascades, our CS. Drafting algorithm has achieved up to 72 percent additional speedup over speculative decoding in our experiments while keeping the same output distribution.

The relationship between communicated language and intended meaning is often probabilistic and sensitive to context. Numerous strategies attempt to estimate such a mapping, often leveraging recursive Bayesian models of communication. In parallel, large language models (LLMs) have been increasingly applied to semantic parsing applications, tasked with inferring logical representations from natural language. While existing LLM explorations have been largely restricted to literal language use, in this work, we evaluate the capacity of LLMs to infer the meanings of pragmatic utterances. Specifically, we explore the case of threshold estimation on the gradable adjective ``strong'', contextually conditioned on a strength prior, then extended to composition with qualification, negation, polarity inversion, and class comparison. We find that LLMs can derive context-grounded, human-like distributions over the interpretations of several complex pragmatic utterances, yet struggle composing with negation. These results inform the inferential capacity of statistical language models, and their use in pragmatic and semantic parsing applications. All corresponding code is made publicly available (https://github.com/benlipkin/probsem/tree/CogSci2023).

Transformer-based language models have been changing the modern Natural Language Processing (NLP) landscape for high-resource languages such as English, Chinese, Russian, etc. However, this technology does not yet exist for any Ghanaian language. In this paper, we introduce the first of such models for Twi or Akan, the most widely spoken Ghanaian language. The specific contribution of this research work is the development of several pretrained transformer language models for the Akuapem and Asante dialects of Twi, paving the way for advances in application areas such as Named Entity Recognition (NER), Neural Machine Translation (NMT), Sentiment Analysis (SA) and Part-of-Speech (POS) tagging. Specifically, we introduce four different flavours of ABENA -- A BERT model Now in Akan that is fine-tuned on a set of Akan corpora, and BAKO - BERT with Akan Knowledge only, which is trained from scratch. We open-source the model through the Hugging Face model hub and demonstrate its use via a simple sentiment classification example.

We are interviewing Walid Saba on *Friday* for Machine Learning Street Talk show (with Yannic Kilcher). He has just written an article, but written many before claiming that deep learning and memorisation / statistical approaches are completely flawed for NLU. He calls these approaches "BERTology" which I think it a funny name! He points out the "the missing text phenomenon" as the biggest issue i.e. "the corner table wants a beer" -- "the _person_ at the corner table wants a beer" ... and provides many other similar examples. He makes a "proof" for this by equating ML to "compressability" and NLU to "expansion" which is intuitive, although I would argue ML could just as easily be used to decompress, think a basic generative model to learn to decompress something.

A language model can predict the probability of the next word in the sequence, based on the words already observed in the sequence. Neural network models are a preferred method for developing statistical language models because they can use a distributed representation where different words with similar meanings have similar representation and because they can use a large context of recently observed words when making predictions. In this tutorial, you will discover how to develop a statistical language model using deep learning in Python. How to Develop a Word-Level Neural Language Model and Use it to Generate Text Photo by Carlo Raso, some rights reserved. The Republic is the classical Greek philosopher Plato's most famous work. It is structured as a dialog (e.g. The entire text is available for free in the public domain. It is available on the Project Gutenberg website in a number of formats. Download the book text and place it in your current working directly with the filename'republic.txt'

Have you ever wondered how Gmail automatic reply works? Or how your phone suggests next word when texting? Or even how a Neural Network can generate musical notes? The general way of generating a sequence of text is to train a model to predict the next word/character given all previous words/characters. Such model is called a Statistical Language Model.

Table-to-text generation aims to generate a description for a factual table which can be viewed as a set of field-value records. To encode both the content and the structure of a table, we propose a novel structure-aware seq2seq architecture which consists of field-gating encoder and description generator with dual attention. In the encoding phase, we update the cell memory of the LSTM unit by a field gate and its corresponding field value in order to incorporate field information into table representation. In the decoding phase, dual attention mechanism which contains word level attention and field level attention is proposed to model the semantic relevance between the generated description and the table. We conduct experiments on the WIKIBIO dataset which contains over 700k biographies and corresponding infoboxes from Wikipedia. The attention visualizations and case studies show that our model is capable of generating coherent and informative descriptions based on the comprehensive understanding of both the content and the structure of a table. Automatic evaluations also show our model outperforms the baselines by a great margin. Code for this work is available on https://github.com/tyliupku/wiki2bio.

We propose a statistical model for natural language that begins by considering language as a monoid, then representing it in complex matrices with a compatible translation invariant probability measure. We interpret the probability measure as arising via the Born rule from a translation invariant matrix product state.