Large Language Models

With Artificial Intelligence (AI) increasingly permeating various aspects of society, including healthcare, the adoption of the Transformers neural network architecture is rapidly changing many applications. Transformer is a type of deep learning architecture initially developed to solve general-purpose Natural Language Processing (NLP) tasks and has subsequently been adapted in many fields, including healthcare. In this survey paper, we provide an overview of how this architecture has been adopted to analyze various forms of healthcare data, including clinical NLP, medical imaging, structured Electronic Health Records (EHR), social media, bio-physiological signals, biomolecular sequences. Furthermore, which have also include the articles that used the transformer architecture for generating surgical instructions and predicting adverse outcomes after surgeries under the umbrella of critical care. Under diverse settings, these models have been used for clinical diagnosis, report generation, data reconstruction, and drug/protein synthesis. Finally, we also discuss the benefits and limitations of using transformers in healthcare and examine issues such as computational cost, model interpretability, fairness, alignment with human values, ethical implications, and environmental impact.

In recent years, AI research has become more and more computationally demanding. In natural language processing (NLP), this tendency is reflected in the emergence of large language models (LLMs) like GPT-3. These powerful neural network-based models can be used for a range of NLP tasks and their language generation capacities have become so sophisticated that it can be very difficult to distinguish their outputs from human language. LLMs have raised concerns over their demonstrable biases, heavy environmental footprints, and future social ramifications. In December 2020, critical research on LLMs led Google to fire Timnit Gebru, co-lead of the company’s AI Ethics team, which sparked a major public controversy around LLMs and the growing corporate influence over AI research. This article explores the role LLMs play in the political economy of AI as infrastructural components for AI research and development. Retracing the technical developments that have led to the emergence of LLMs, we point out how they are intertwined with the business model of big tech companies and further shift power relations in their favour. This becomes visible through the Transformer, which is the underlying architecture of most LLMs today and started the race for ever bigger models when it was introduced by Google in 2017. Using the example of GPT-3, we shed light on recent corporate efforts to commodify LLMs through paid API access and exclusive licensing, raising questions around monopolization and dependency in a field that is increasingly divided by access to large-scale computing power.

In this work, we propose Retentive Network (RetNet) as a foundation architecture for large language models, simultaneously achieving training parallelism, low-cost inference, and good performance. We theoretically derive the connection between recurrence and attention. Then we propose the retention mechanism for sequence modeling, which supports three computation paradigms, i.e., parallel, recurrent, and chunkwise recurrent. Specifically, the parallel representation allows for training parallelism. The recurrent representation enables low-cost $O(1)$ inference, which improves decoding throughput, latency, and GPU memory without sacrificing performance. The chunkwise recurrent representation facilitates efficient long-sequence modeling with linear complexity, where each chunk is encoded parallelly while recurrently summarizing the chunks. Experimental results on language modeling show that RetNet achieves favorable scaling results, parallel training, low-cost deployment, and efficient inference. The intriguing properties make RetNet a strong successor to Transformer for large language models. Code will be available at https://aka.ms/retnet.

Transformers, originally introduced for machine translation, and built upon the Self-Attention mechanism, have undergone a remarkable evolution, establishing themselves as the bedrock of large language models (LLMs). Their unparalleled capacity to model intricate relationships and capture extensive dependencies within sequences has propelled their prominence. This article, presented in a popular science format, serves as an introduction to the transformer architecture, elucidating its innovative structure that enables efficient processing of long sequences and capturing dependencies over extended distances. We believe that this resource will prove valuable to college students or youth researchers aspiring to delve into the study and research of modern Artificial Intelligence (AI) domains.

In the rapidly evolving landscape of genomics, deep learning has emerged as a useful tool for tackling complex computational challenges. This review focuses on the transformative role of Large Language Models (LLMs), which are mostly based on the transformer architecture, in genomics. Building on the foundation of traditional convolutional neural networks and recurrent neural networks, we explore both the strengths and limitations of transformers and other LLMs for genomics. Additionally, we contemplate the future of genomic modeling beyond the transformer architecture based on current trends in research. The paper aims to serve as a guide for computational biologists and computer scientists interested in LLMs for genomic data. We hope the paper can also serve as an educational introduction and discussion for biologists to a fundamental shift in how we will be analyzing genomic data in the future.

We will introduce Large Language Models (LLMs), focusing on their core architecture, training methodologies, practical applications, and some important considerations. We begin with a brief exploration of the transformer architecture, examining how architectural choices and scaling laws have shaped modern language models. We then investigate key training paradigms, with particular emphasis on Reinforcement Learning from Human Feedback (RLHF) and instruction tuning, while briefly touching on emerging approaches like Direct Preference Optimisation (DPO). The discussion progresses through the evolution of knowledge integration techniques, from basic context stuffing to Retrieval-Augmented Generation (RAG), highlighting their practical implications. We will discuss the critical issue of hallucinations in LLM outputs, providing concrete examples and verification strategies. To conclude, a quick overview of the current software landscape and immediate future developments in the field, and a look ahead to the rest of the webinar series.

Generative AI has transitioned from the lab to the classroom, hospitals, software teams, and boardrooms. Large language models (LLMs), transformer-based foundation models, are at the heart of this paradigm change, with their ability to write texts, reason, summarize, translate, and code. However, these developments have left many students and researchers, as well as practitioners, with disjointed bits of knowledge. This book was developed with that knowledge in mind in order to streamline all that information in a single comprehensive and hands-on natural language processing (NLP) and generative AI guide. The book follows the path of the technology itself. Chapter 1 covers transformer architectures ranging from the self-attention mechanism to encoder-decoder and decoder-only architectures. The course also includes lectures on mixture of experts, scaling laws, pretraining, fine-tuning, alignment and multimodal models. Targeting prompt engineering and in-context learning is the focus of Chapter 2. It uses zero shot, few shot, chain of thought prompting, system prompts and context engineering as skills to learn and test. Chapter 3 focuses on LLM reasoning and test-time compute, presenting various ways in which deep learning systems can be extended to form reasoning models, self-consistency, and search over reasoning paths. As it is not possible to know everything, Chapters 4 and 5 discuss retrieval-augmented generation (RAG). They start with embeddings and vector databases, chunking strategies and semantic search. They then transition to advanced RAG – hybrid search, reranking, knowledge graphs, GraphRAG, and agentic RAG, all while also focusing on evaluation and LLMOps for production. The strands are woven together in Chapter 6 under the umbrella of agentic AI and AI agents. It examines the use of tools, the calling of functions, the memory of an agent, as well as planning and multi-agent systems and the way autonomous agents are transforming work. Chapter 7 ends with trustworthy and responsible AI. It speaks about hallucination, bias/fairness, explainability, data privacy, AI safety, AI ethics and AI governance, as without trust, the capability cannot be sustained. The chapters are designed with learning and reference in mind for each. It introduces the foundations and outlines the key techniques side-by-side in summary tables, covers present open challenges and future directions, and concludes with practical advice. Benchmarking and evaluation is handled very transparently with regard to the deficiencies in existing approaches – and I do hope readers will take the shortcomings as a challenge and a starting point for their own work. The working levels that the book is aimed at are graduate and undergraduate students, academic researchers, machine learning engineers, data scientists, and technology leaders. Can also be used as a textbook in the course where it was used or as a builder's handbook or as a literature map for those preparing an indexed study in a Scopus journal. I'm focusing more on enduring principles than fleeting tools because no one can stay current in a science that is evolving monthly. We hope this book empowers you to grasp generative AI, create responsibly, and shape the future of AI together.

Financial text is produced and interpreted within a market environment, yet financial text classifiers almost always receive text alone. We study whether financial time series are useful as an additional input on the task of classifying sentences from Federal Reserve communication as hawkish, dovish, or neutral. Our system, \lfts{}, extends the \lf{} architecture with this modality: a small voting network combines three independently trained components, a fine-tuned RoBERTa encoder, a prompted large language model (LLM), and a fused ensemble of time-series transformers over the market series of the months preceding publication. Because only about a thousand annotated sentences are available for training, the RoBERTa encoder is first pre-trained on sentences annotated automatically by the LLM and only then fine-tuned on the human labels. Trained on Federal Open Market Committee (FOMC) communication up to 2015 and evaluated on 2015--2022, the fused system achieves 70.2\% weighted F1 -- against 64.1\% for the zero-shot LLM -- and overtakes it with as few as 240 human-labelled sentences. We take this as initial evidence for market time series as an input modality in financial text classification.