Llm-Adapters: An Adapter Family For Parameter-Efficient Fine-Tuning Of Large Language Models,
2023
Singapore Management University
Llm-Adapters: An Adapter Family For Parameter-Efficient Fine-Tuning Of Large Language Models, Zhiqiang Hu, Lei Wang, Yihuai Lan, Wanyu Xu, Ee-Peng Lim, Lidong Bing, Xing Xu, Soujanya Poria, Roy Ka-Wei Lee
Research Collection School Of Computing and Information Systems
The success of large language models (LLMs), like GPT-4 and ChatGPT, has led to the development of numerous cost-effective and accessible alternatives that are created by finetuning open-access LLMs with task-specific data (e.g., ChatDoctor) or instruction data (e.g., Alpaca). Among the various fine-tuning methods, adapter-based parameter-efficient fine-tuning (PEFT) is undoubtedly one of the most attractive topics, as it only requires fine-tuning a few external parameters instead of the entire LLMs while achieving comparable or even better performance. To enable further research on PEFT methods of LLMs, this paper presents LLMAdapters, an easy-to-use framework that integrates various adapters into LLMs and …
Graph Contrastive Learning With Stable And Scalable Spectral Encoding,
2023
Singapore Management University
Graph Contrastive Learning With Stable And Scalable Spectral Encoding, Deyu Bo, Yuan Fang, Yang Liu, Chuan Shi
Research Collection School Of Computing and Information Systems
Graph contrastive learning (GCL) aims to learn representations by capturing the agreements between different graph views. Traditional GCL methods generate views in the spatial domain, but it has been recently discovered that the spectral domain also plays a vital role in complementing spatial views. However, existing spectral-based graph views either ignore the eigenvectors that encode valuable positional information, or suffer from high complexity when trying to address the instability of spectral features. To tackle these challenges, we first design an informative, stable, and scalable spectral encoder, termed EigenMLP, to learn effective representations from the spectral features. Theoretically, EigenMLP is invariant …
Examining The Inter-Consistency Of Large Language Models: An In-Depth Analysis Via Debate,
2023
Singapore Management University
Examining The Inter-Consistency Of Large Language Models: An In-Depth Analysis Via Debate, Kai Xiong, Xiao Ding, Yixin Cao, Ting Liu, Bing Qin
Research Collection School Of Computing and Information Systems
Large Language Models (LLMs) have shown impressive capabilities in various applications, but they still face various inconsistency issues. Existing works primarily focus on the inconsistency issues within a single LLM, while we complementarily explore the inter-consistency among multiple LLMs for collaboration. To examine whether LLMs can collaborate effectively to achieve a consensus for a shared goal, we focus on commonsense reasoning, and introduce a formal debate framework (FORD) to conduct a three-stage debate among LLMs with real-world scenarios alignment: fair debate, mismatched debate, and roundtable debate. Through extensive experiments on various datasets, LLMs can effectively collaborate to reach a consensus …
Learning To Search Feasible And Infeasible Regions Of Routing Problems With Flexible Neural K-Opt,
2023
Singapore Management University
Learning To Search Feasible And Infeasible Regions Of Routing Problems With Flexible Neural K-Opt, Yining Ma, Zhiguang Cao, Yew Meng Chee
Research Collection School Of Computing and Information Systems
In this paper, we present Neural k-Opt (NeuOpt), a novel learning-to-search (L2S) solver for routing problems. It learns to perform flexible k-opt exchanges based on a tailored action factorization method and a customized recurrent dual-stream decoder. As a pioneering work to circumvent the pure feasibility masking scheme and enable the autonomous exploration of both feasible and infeasible regions, we then propose the Guided Infeasible Region Exploration (GIRE) scheme, which supplements the NeuOpt policy network with feasibility-related features and leverages reward shaping to steer reinforcement learning more effectively. Besides, we further equip NeuOpt with dynamic data augmentations during inference for more …
Exgen: Ready-To-Use Exercise Generation In Introductory Programming Courses,
2023
Singapore Management University
Exgen: Ready-To-Use Exercise Generation In Introductory Programming Courses, Nguyen Binh Duong Ta, Hua Gia Phuc Nguyen, Gottipati Swapna
Research Collection School Of Computing and Information Systems
In introductory programming courses, students as novice programmers would benefit from doing frequent practices set at a difficulty level and concept suitable for their skills and knowledge. However, setting many good programming exercises for individual learners is very time-consuming for instructors. In this work, we propose an automated exercise generation system, named ExGen, which leverages recent advances in pre-trained large language models (LLMs) to automatically create customized and ready-to-use programming exercises for individual students ondemand. The system integrates seamlessly with Visual Studio Code, a popular development environment for computing students and software engineers. ExGen effectively does the following: 1) maintaining …
Spatial-Temporal Episodic Memory Modeling For Adls: Encoding, Retrieval, And Prediction,
2023
China University of Petroleum
Spatial-Temporal Episodic Memory Modeling For Adls: Encoding, Retrieval, And Prediction, Xinjing Song, Di Wang, Chai Quek, Ah-Hwee Tan, Yanjiang Wang
Research Collection School Of Computing and Information Systems
Activities of daily living (ADLs) relate to people’s daily self-care activities, which reflect their living habits and lifestyle. A prior study presented a neural network model called STADLART for ADL routine learning. In this paper, we propose a cognitive model named Spatial-Temporal Episodic Memory for ADL (STEM-ADL), which extends STADLART to encode event sequences in the form of distributed episodic memory patterns. Specifically, STEM-ADL encodes each ADL and its associated contextual information as an event pattern and encodes all events in a day as an episode pattern. By explicitly encoding the temporal characteristics of events as activity gradient patterns, STEM-ADL …
Generalized Logit Adjustment: Calibrating Fine-Tuned Models By Removing Label Bias In Foundation Models,
2023
Singapore Management University
Generalized Logit Adjustment: Calibrating Fine-Tuned Models By Removing Label Bias In Foundation Models, Beier Zhu, Kaihua Tang, Qianru Sun, Hanwang Zhang
Research Collection School Of Computing and Information Systems
Foundation models like CLIP allow zero-shot transfer on various tasks without additional training data. Yet, the zero-shot performance is less competitive than a fully supervised one. Thus, to enhance the performance, fine-tuning and ensembling are also commonly adopted to better fit the downstream tasks. However, we argue that such prior work has overlooked the inherent biases in foundation models. Due to the highly imbalanced Web-scale training set, these foundation models are inevitably skewed toward frequent semantics, and thus the subsequent fine-tuning or ensembling is still biased. In this study, we systematically examine the biases in foundation models and demonstrate the …
Flowpg: Action-Constrained Policy Gradient With Normalizing Flows,
2023
Singapore Management University
Flowpg: Action-Constrained Policy Gradient With Normalizing Flows, Brahmanage Janaka Chathuranga Thilakarathna, Jiajing Ling, Akshat Kumar
Research Collection School Of Computing and Information Systems
Action-constrained reinforcement learning (ACRL) is a popular approach for solving safety-critical and resource-allocation related decision making problems. A major challenge in ACRL is to ensure agent taking a valid action satisfying constraints in each RL step. Commonly used approach of using a projection layer on top of the policy network requires solving an optimization program which can result in longer training time, slow convergence, and zero gradient problem. To address this, first we use a normalizing flow model to learn an invertible, differentiable mapping between the feasible action space and the support of a simple distribution on a latent variable, …
On The Usage Of Continual Learning For Out-Of-Distribution Generalization In Pre-Trained Language Models Of Code,
2023
Singapore Management University
On The Usage Of Continual Learning For Out-Of-Distribution Generalization In Pre-Trained Language Models Of Code, Martin Weyssow, Xin Zhou, Kisub Kim, David Lo, Houari A. Sahraoui
Research Collection School Of Computing and Information Systems
Pre-trained language models (PLMs) have become a prevalent technique in deep learning for code, utilizing a two-stage pre-training and fine-tuning procedure to acquire general knowledge about code and specialize in a variety of downstream tasks. However, the dynamic nature of software codebases poses a challenge to the effectiveness and robustness of PLMs. In particular, world-realistic scenarios potentially lead to significant differences between the distribution of the pre-training and test data, i.e., distribution shift, resulting in a degradation of the PLM's performance on downstream tasks. In this paper, we stress the need for adapting PLMs of code to software data whose …
End-To-End Task-Oriented Dialogue: A Survey Of Tasks, Methods, And Future Directions,
2023
Central South University
End-To-End Task-Oriented Dialogue: A Survey Of Tasks, Methods, And Future Directions, Libo Qin, Wenbo Pan, Qiguang Chen, Lizi Liao, Zhou Yu, Yue Zhang, Wanxiang Che, Min Li
Research Collection School Of Computing and Information Systems
End-to-end task-oriented dialogue (EToD) can directly generate responses in an end-to-end fashion without modular training, which attracts escalating popularity. The advancement of deep neural networks, especially the successful use of large pre-trained models, has further led to significant progress in EToD research in recent years. In this paper, we present a thorough review and provide a unified perspective to summarize existing approaches as well as recent trends to advance the development of EToD research. The contributions of this paper can be summarized: (1) First survey: to our knowledge, we take the first step to present a thorough survey of this …
Generative Modelling Of Stochastic Actions With Arbitrary Constraints In Reinforcement Learning,
2023
Singapore Management University
Generative Modelling Of Stochastic Actions With Arbitrary Constraints In Reinforcement Learning, Changyu Chen, Ramesha Karunasena, Thanh Hong Nguyen, Arunesh Sinha, Pradeep Varakantham
Research Collection School Of Computing and Information Systems
Many problems in Reinforcement Learning (RL) seek an optimal policy with large discrete multidimensional yet unordered action spaces; these include problems in randomized allocation of resources such as placements of multiple security resources and emergency response units, etc. A challenge in this setting is that the underlying action space is categorical (discrete and unordered) and large, for which existing RL methods do not perform well. Moreover, these problems require validity of the realized action (allocation); this validity constraint is often difficult to express compactly in a closed mathematical form. The allocation nature of the problem also prefers stochastic optimal policies, …
Beyond Factuality: A Comprehensive Evaluation Of Large Language Models As Knowledge Generators,
2023
Singapore Management University
Beyond Factuality: A Comprehensive Evaluation Of Large Language Models As Knowledge Generators, Liang Chen, Yang Deng, Yatao Bian, Zeyu Qin, Bingzhe Wu, Tat-Seng Chua, Kam-Fai Wong
Research Collection School Of Computing and Information Systems
Large language models (LLMs) outperform information retrieval techniques for downstream knowledge-intensive tasks when being prompted to generate world knowledge. Yet, community concerns abound regarding the factuality and potential implications of using this uncensored knowledge. In light of this, we introduce CONNER, a COmpreheNsive kNowledge Evaluation fRamework, designed to systematically and automatically evaluate generated knowledge from six important perspectives - Factuality, Relevance, Coherence, Informativeness, Helpfulness and Validity. We conduct an extensive empirical analysis of the generated knowledge from three different types of LLMs on two widely-studied knowledge-intensive tasks, i.e., open-domain question answering and knowledge-grounded dialogue. Surprisingly, our study reveals that the …
Cue-Cot: Chain-Of-Thought Prompting For Responding To In-Depth Dialogue Questions With Llms,
2023
Singapore Management University
Cue-Cot: Chain-Of-Thought Prompting For Responding To In-Depth Dialogue Questions With Llms, Hongru Wang, Rui Wang, Fei Mi, Yang Deng, Zezhong Wang, Bin Liang, Ruifeng Xu, Kam-Fai Wong
Research Collection School Of Computing and Information Systems
Large Language Models (LLMs), such as ChatGPT, greatly empower dialogue systems with strong language understanding and generation capabilities. However, most of the previous works prompt the LLMs to directly generate a response based on the dialogue context, overlooking the underlying linguistic cues about the user status exhibited in the context. Such in-depth dialogue scenarios are challenging for existing LLMs to figure out the user’s hidden needs and respond satisfactorily through a single-step inference. To this end, we propose a novel linguistic cue-based chain-of-thoughts (Cue-CoT), which enhances the LLMs inference with an intermediate reasoning step to find cues exhibited in the …
Localization Of Gamma-Ray Bursts In A Balloon-Borne Telescope,
2023
Missouri University of Science and Technology
Localization Of Gamma-Ray Bursts In A Balloon-Borne Telescope, Ye Htet, Marion Sudvarg, Jeremy Buhler, Roger Chamberlain, James Buckley
Computer Science Faculty Research & Creative Works
Multi-messenger astrophysics combines observations from multiple instruments to study transient astrophysical phenomena, many occurring at seconds-level timescales. To identify and precisely localize these events in the sky, current systems often search through extensive sensor data, requiring resource-intensive computation to achieve results on the timescale of the events themselves. We seek to reduce computational requirements so as to perform real-time event localization with limited computational resources suitable for an orbital platform. This work studies the performance of a computational pipeline for real-time gamma-ray burst (GRB) detection and localization aboard the Antarctic Demonstrator for the Advanced Particle-astro-physics Telescope (ADAPT), a balloon-borne prototype …
Uavs And Deep Neural Networks: An Alternative Approach To Monitoring Waterfowl At The Site Level,
2023
Louisiana State University and Agricultural and Mechanical College
Uavs And Deep Neural Networks: An Alternative Approach To Monitoring Waterfowl At The Site Level, Zachary J. Loken
LSU Master's Theses
Understanding how waterfowl respond to habitat restoration and management activities is crucial for evaluating and refining conservation delivery programs. However, site-specific waterfowl monitoring is challenging, especially in heavily forested systems such as the Mississippi Alluvial Valley (MAV)—a primary wintering region for ducks in North America. I hypothesized that using uncrewed aerial vehicles (UAVs) coupled with deep learning-based methods for object detection would provide an efficient and effective means for surveying non-breeding waterfowl on difficult-to-access restored wetland sites. Accordingly, during the winters of 2021 and 2022, I surveyed wetland restoration easements in the MAV using a UAV equipped with a dual …
All Quiet On The Digital Front: The Unseen Psychological Impacts On Cybersecurity First Responders,
2023
University of South Florida
All Quiet On The Digital Front: The Unseen Psychological Impacts On Cybersecurity First Responders, Tammie R. Hollis
USF Tampa Graduate Theses and Dissertations
Driven by the increasing frequency of cyberattacks and the existing talent gap between industry needs and skilled professionals, this research study focused on the crucial human element in the domain of cybersecurity incident response. The objective of this dissertation was to offer a meaningful exploration of the lived experiences encountered by cybersecurity incident responders and an assessment of the subsequent impacts on their well-being. Additionally, this study sought to draw comparisons between the experiences of cybersecurity incident responders and their counterparts in traditional emergency response roles. Semi-structured interviews were conducted with a cohort of 22 individuals with first-hand experience working …
Modelling Prediction Of Cities Real Estate Price Trend Using Recurrent Neural Network: A Case Of Dar Es Salaam City,
2023
1Department of Computer Science and Engineering, College of Information and Communication Technologies, University of Dar es Salaam
Modelling Prediction Of Cities Real Estate Price Trend Using Recurrent Neural Network: A Case Of Dar Es Salaam City, Ellen Kalinga
Tanzania Journal of Engineering and Technology (TJET)
Real estate refers to a class of real property such as land and its associated infrastructure. The prediction of real estate prices in cities, which is affected by a number of parameters, is an open research problem. The lack of reliable and effective tools for price forecasting in real estate, especially in residential housing, can adversely affect investment flows and the growth of the real estate sector. Taking Tanzania as an example, the price prediction practices rely on human suggestions that are prone to personal bias and subjective to price hysteria for personal gain and impact consumer expectations. To address …
Bridging Domain Gaps For Cross-Spectrum And Long-Range Face Recognition Using Domain Adaptive Machine Learning,
2023
University of Nebraska-Lincoln
Bridging Domain Gaps For Cross-Spectrum And Long-Range Face Recognition Using Domain Adaptive Machine Learning, Cedric Armel Nimpa Fondje
Dissertations and Doctoral Documents, University of Nebraska-Lincoln, 2023–
Face recognition technology has witnessed significant advancements in recent decades, enabling its widespread adoption in various applications such as security, surveillance, and biometrics applications. However, one of the primary challenges faced by existing face recognition systems is their limited performance when presented with images from different modalities or domains( such as infrared to visible, long range to close range, nighttime to daytime, profile to f rontal, etc.) Additionally, advancements in camera sensors, analytics beyond the visible spectrum, and the increasing size of cross-modal datasets have led to a particular interest in cross-modal learning for face recognition in the biometrics and …
Spatial Data Management For Green Mobility,
2023
Singapore Management University
Spatial Data Management For Green Mobility, Christophe Claramunt, Christine Bassem, Demetrios Zeinalipour-Yazti, Baihua Zheng, Goce Trajcevski, Kristian Torp
Research Collection School Of Computing and Information Systems
While many countries are developing appropriate actions towards a greener future and moving towards adopting sustainable mobility activities, the real-time management and planning of innovative transportation facilities and services in urban environments still require the development of advanced mobile data management infrastructures. Novel green mobility solutions, such as electric, hybrid, solar and hydrogen vehicles, as well as public and gig-based transportation resources are very likely to reduce the carbon footprint. However, their successful implementation still needs efficient spatio-temporal data management resources and applications to provide a clear picture and demonstrate their effectiveness. This paper discusses the major data management challenges, …
Leveraging Long Short-Term User Preference In Conversational Recommendation Via Multi-Agent Reinforcement Learning,
2023
Singapore Management University
Leveraging Long Short-Term User Preference In Conversational Recommendation Via Multi-Agent Reinforcement Learning, Yang Deng, Yaliang Li, Bolin Ding, Wai Lam
Research Collection School Of Computing and Information Systems
Conversational recommender systems (CRS) endow traditional recommender systems with the capability of dynamically obtaining users’ short-term preferences for items and attributes through interactive dialogues. There are three core challenges for CRS, including the intelligent decisions for what attributes to ask, which items to recommend, and when to askor recommend, at each conversation turn. Previous methods mainly leverage reinforcement learning (RL) to learn conversational recommendation policies for solving one or two of these three decision-making problems in CRS with separated conversation and recommendation components. These approaches restrict the scalability and generality of CRS and fall short of preserving a stable training …
