Jules White

Jules White, Quchen Fu, Sam Hays et al.

Prompt engineering is an increasingly important skill set needed to converse effectively with large language models (LLMs), such as ChatGPT. Prompts are instructions given to an LLM to enforce rules, automate processes, and ensure specific qualities (and quantities) of generated output. Prompts are also a form of programming that can customize the outputs and interactions with an LLM. This paper describes a catalog of prompt engineering techniques presented in pattern form that have been applied to solve common problems when conversing with LLMs. Prompt patterns are a knowledge transfer method analogous to software patterns since they provide reusable solutions to common problems faced in a particular context, i.e., output generation and interaction when working with LLMs. This paper provides the following contributions to research on prompt engineering that apply LLMs to automate software development tasks. First, it provides a framework for documenting patterns for structuring prompts to solve a range of problems so that they can be adapted to different domains. Second, it presents a catalog of patterns that have been applied successfully to improve the outputs of LLM conversations. Third, it explains how prompts can be built from multiple patterns and illustrates prompt patterns that benefit from combination with other prompt patterns.

Jules White, Sam Hays, Quchen Fu et al.

This paper presents prompt design techniques for software engineering, in the form of patterns, to solve common problems when using large language models (LLMs), such as ChatGPT to automate common software engineering activities, such as ensuring code is decoupled from third-party libraries and simulating a web application API before it is implemented. This paper provides two contributions to research on using LLMs for software engineering. First, it provides a catalog of patterns for software engineering that classifies patterns according to the types of problems they solve. Second, it explores several prompt patterns that have been applied to improve requirements elicitation, rapid prototyping, code quality, refactoring, and system design.

Henry Gilbert, Michael Sandborn, Douglas C. Schmidt et al.

The rise of large language models (LLMs) is revolutionizing information retrieval, question answering, summarization, and code generation tasks. However, in addition to confidently presenting factually inaccurate information at times (known as "hallucinations"), LLMs are also inherently limited by the number of input and output tokens that can be processed at once, making them potentially less effective on tasks that require processing a large set or continuous stream of information. A common approach to reducing the size of data is through lossless or lossy compression. Yet, in some cases it may not be strictly necessary to perfectly recover every detail from the original data, as long as a requisite level of semantic precision or intent is conveyed. This paper presents three contributions to research on LLMs. First, we present the results from experiments exploring the viability of approximate compression using LLMs, focusing specifically on GPT-3.5 and GPT-4 via ChatGPT interfaces. Second, we investigate and quantify the capability of LLMs to compress text and code, as well as to recall and manipulate compressed representations of prompts. Third, we present two novel metrics -- Exact Reconstructive Effectiveness (ERE) and Semantic Reconstruction Effectiveness (SRE) -- that quantify the level of preserved intent between text compressed and decompressed by the LLMs we studied. Our initial results indicate that GPT-4 can effectively compress and reconstruct text while preserving the semantic essence of the original text, providing a path to leverage $\sim$5$\times$ more tokens than present limits allow.

Quchen Fu, Zhongwei Teng, Marco Georgaklis et al.

Translating natural language into Bash Commands is an emerging research field that has gained attention in recent years. Most efforts have focused on producing more accurate translation models. To the best of our knowledge, only two datasets are available, with one based on the other. Both datasets involve scraping through known data sources (through platforms like stack overflow, crowdsourcing, etc.) and hiring experts to validate and correct either the English text or Bash Commands. This paper provides two contributions to research on synthesizing Bash Commands from scratch. First, we describe a state-of-the-art translation model used to generate Bash Commands from the corresponding English text. Second, we introduce a new NL2CMD dataset that is automatically generated, involves minimal human intervention, and is over six times larger than prior datasets. Since the generation pipeline does not rely on existing Bash Commands, the distribution and types of commands can be custom adjusted. We evaluate the performance of ChatGPT on this task and discuss the potential of using it as a data generator. Our empirical results show how the scale and diversity of our dataset can offer unique opportunities for semantic parsing researchers.

Quchen Fu, Ramesh Chukka, Keith Achorn et al.

GPUs have been favored for training deep learning models due to their highly parallelized architecture. As a result, most studies on training optimization focus on GPUs. There is often a trade-off, however, between cost and efficiency when deciding on how to choose the proper hardware for training. In particular, CPU servers can be beneficial if training on CPUs was more efficient, as they incur fewer hardware update costs and better utilizing existing infrastructure. This paper makes several contributions to research on training deep learning models using CPUs. First, it presents a method for optimizing the training of deep learning models on Intel CPUs and a toolkit called ProfileDNN, which we developed to improve performance profiling. Second, we describe a generic training optimization method that guides our workflow and explores several case studies where we identified performance issues and then optimized the Intel Extension for PyTorch, resulting in an overall 2x training performance increase for the RetinaNet-ResNext50 model. Third, we show how to leverage the visualization capabilities of ProfileDNN, which enabled us to pinpoint bottlenecks and create a custom focal loss kernel that was two times faster than the official reference PyTorch implementation.

Sander Schulhoff, Michael Ilie, Nishant Balepur et al.

Generative Artificial Intelligence (GenAI) systems are increasingly being deployed across diverse industries and research domains. Developers and end-users interact with these systems through the use of prompting and prompt engineering. Although prompt engineering is a widely adopted and extensively researched area, it suffers from conflicting terminology and a fragmented ontological understanding of what constitutes an effective prompt due to its relatively recent emergence. We establish a structured understanding of prompt engineering by assembling a taxonomy of prompting techniques and analyzing their applications. We present a detailed vocabulary of 33 vocabulary terms, a taxonomy of 58 LLM prompting techniques, and 40 techniques for other modalities. Additionally, we provide best practices and guidelines for prompt engineering, including advice for prompting state-of-the-art (SOTA) LLMs such as ChatGPT. We further present a meta-analysis of the entire literature on natural language prefix-prompting. As a culmination of these efforts, this paper presents the most comprehensive survey on prompt engineering to date.

Zhongwei Teng, Quchen Fu, Jules White et al.

An emerging trend in audio processing is capturing low-level speech representations from raw waveforms. These representations have shown promising results on a variety of tasks, such as speech recognition and speech separation. Compared to handcrafted features, learning speech features via backpropagation provides the model greater flexibility in how it represents data for different tasks theoretically. However, results from empirical study shows that, in some tasks, such as voice spoof detection, handcrafted features are more competitive than learned features. Instead of evaluating handcrafted features and raw waveforms independently, this paper proposes an Auxiliary Rawnet model to complement handcrafted features with features learned from raw waveforms. A key benefit of the approach is that it can improve accuracy at a relatively low computational cost. The proposed Auxiliary Rawnet model is tested using the ASVspoof 2019 dataset and the results from this dataset indicate that a light-weight waveform encoder can potentially boost the performance of handcrafted-features-based encoders in exchange for a small amount of additional computational work.

Quchen Fu, Zhongwei Teng, Jules White et al.

Voice assistants, such as smart speakers, have exploded in popularity. It is currently estimated that the smart speaker adoption rate has exceeded 35% in the US adult population. Manufacturers have integrated speaker identification technology, which attempts to determine the identity of the person speaking, to provide personalized services to different members of the same family. Speaker identification can also play an important role in controlling how the smart speaker is used. For example, it is not critical to correctly identify the user when playing music. However, when reading the user's email out loud, it is critical to correctly verify the speaker that making the request is the authorized user. Speaker verification systems, which authenticate the speaker identity, are therefore needed as a gatekeeper to protect against various spoofing attacks that aim to impersonate the enrolled user. This paper compares popular learnable front-ends which learn the representations of audio by joint training with downstream tasks (End-to-End). We categorize the front-ends by defining two generic architectures and then analyze the filtering stages of both types in terms of learning constraints. We propose replacing fixed filterbanks with a learnable layer that can better adapt to anti-spoofing tasks. The proposed FastAudio front-end is then tested with two popular back-ends to measure the performance on the LA track of the ASVspoof 2019 dataset. The FastAudio front-end achieves a relative improvement of 27% when compared with fixed front-ends, outperforming all other learnable front-ends on this task.

Mayank Agarwal, Tathagata Chakraborti, Quchen Fu et al.

The NLC2CMD Competition hosted at NeurIPS 2020 aimed to bring the power of natural language processing to the command line. Participants were tasked with building models that can transform descriptions of command line tasks in English to their Bash syntax. This is a report on the competition with details of the task, metrics, data, attempted solutions, and lessons learned.

Peng Zhang, Douglas C. Schmidt, Jules White

Known for its decentralized and tamper-aware properties, blockchain is attractive to enhance the infrastructure of systems that have been constrained by traditionally centralized and vendor-locked environments. Although blockchain has commonly been used as the operational model behind cryptocurrency, it has far more foreseeable utilities in domains like healthcare, where efficient data flow is highly demanded. Particularly, blockchain and related technologies have been touted as foundational technologies for addressing healthcare interoperability challenges, such as promoting effective communications and securing data exchanges across various healthcare systems. Despite the increasing interests in leveraging blockchain technology to improve healthcare infrastructures, a major gap in literature is the lack of available recommendations for concrete architectural styles and design considerations for creating blockchain-based apps and systems with a healthcare focus. This research provides two contributions to bridge the gap in existing research. First, we introduce a pattern sequence for designing blockchain-based healthcare systems focused on secure and at-scale data exchange. Our approach adapts traditional software patterns and proposes novel patterns that take into account both the technical requirements specific to healthcare systems and the implications of these requirements on naive blockchain-based solutions. Second, we provide a pattern-oriented reference architecture using an example application of the pattern sequence for guiding software developers to design interoperable (on the technical level) healthcare IT systems atop blockchain-based infrastructures. The reference architecture focuses on minimizing storage requirements on-chain, preserving the privacy of sensitive information, facilitating scalable communications, and maximizing evolvability of the system.

Peng Zhang, Jules White, Douglas C. Schmidt et al.

Secure and scalable data sharing is essential for collaborative clinical decision making. Conventional clinical data efforts are often siloed, however, which creates barriers to efficient information exchange and impedes effective treatment decision made for patients. This paper provides four contributions to the study of applying blockchain technology to clinical data sharing in the context of technical requirements defined in the "Shared Nationwide Interoperability Roadmap" from the Office of the National Coordinator for Health Information Technology (ONC). First, we analyze the ONC requirements and their implications for blockchain-based systems. Second, we present FHIRChain, which is a blockchain-based architecture designed to meet ONC requirements by encapsulating the HL7 Fast Healthcare Interoperability Resources (FHIR) standard for shared clinical data. Third, we demonstrate a FHIRChain-based decentralized app using digital health identities to authenticate participants in a case study of collaborative decision making for remote cancer care. Fourth, we highlight key lessons learned from our case study.

Alex Cheng, Jules White

Outpatient clinics often run behind schedule due to patients who arrive late or appointments that run longer than expected. We sought to develop a generalizable method that would allow healthcare providers to diagnose problems in workflow that disrupt the schedule on any given provider clinic day. We use a constraint optimization problem to identify the least number of appointment modifications that make the rest of the schedule run on-time. We apply this method to an outpatient clinic at Vanderbilt. For patient seen in this clinic between March 27, 2017 and April 21, 2017, long cycle times tended to affect the overall schedule more than late patients. Results from this workflow diagnosis method could be used to inform interventions to help clinics run smoothly, thus decreasing patient wait times and increasing provider utilization.

Fangzhou sun, Abhishek Dubey, Jules White

Non-recurring traffic congestion is caused by temporary disruptions, such as accidents, sports games, adverse weather, etc. We use data related to real-time traffic speed, jam factors (a traffic congestion indicator), and events collected over a year from Nashville, TN to train a multi-layered deep neural network. The traffic dataset contains over 900 million data records. The network is thereafter used to classify the real-time data and identify anomalous operations. Compared with traditional approaches of using statistical or machine learning techniques, our model reaches an accuracy of 98.73 percent when identifying traffic congestion caused by football games. Our approach first encodes the traffic across a region as a scaled image. After that the image data from different timestamps is fused with event- and time-related data. Then a crossover operator is used as a data augmentation method to generate training datasets with more balanced classes. Finally, we use the receiver operating characteristic (ROC) analysis to tune the sensitivity of the classifier. We present the analysis of the training time and the inference time separately.