arXiv subjects

Kevin Zhu

Publications and source records attributed to Kevin Zhu.

126 records · Page 7Linked to original sources

arXiv2024

A Comparative Study of Translation Bias and Accuracy in Multilingual Large Language Models for Cross-Language Claim Verification

The rise of digital misinformation has heightened interest in using multilingual Large Language Models (LLMs) for fact-checking. This study systematically evaluates translation bias and the effectiveness of LLMs for cross-lingual claim verification across 15 languages from five language families: Romance, Slavic, Turkic, Indo-Aryan, and Kartvelian. Using the XFACT dataset to assess their impact on accuracy and bias, we investigate two distinct translation methods: pre-translation and self-translation. We use mBERT's performance on the English dataset as a baseline to compare language-specific accuracies. Our findings reveal that low-resource languages exhibit significantly lower accuracy in direct inference due to underrepresentation in the training data. Furthermore, larger models demonstrate superior performance in self-translation, improving translation accuracy and reducing bias. These results highlight the need for balanced multilingual training, especially in low-resource languages, to promote equitable access to reliable fact-checking tools and minimize the risk of spreading misinformation in different linguistic contexts.

cs.CL↗

arXiv2024

Fine-Tuning Language Models for Ethical Ambiguity: A Comparative Study of Alignment with Human Responses

Language models often misinterpret human intentions due to their handling of ambiguity, a limitation well-recognized in NLP research. While morally clear scenarios are more discernible to LLMs, greater difficulty is encountered in morally ambiguous contexts. In this investigation, we explored LLM calibration to show that human and LLM judgments are poorly aligned in such scenarios. We used two curated datasets from the Scruples project for evaluation: DILEMMAS, which involves pairs of distinct moral scenarios to assess the model's ability to compare and contrast ethical situations, and ANECDOTES, which presents individual narratives to evaluate the model's skill in drawing out details, interpreting, and analyzing distinct moral scenarios. Model answer probabilities were extracted for all possible choices and compared with human annotations to benchmark the alignment of three models: Llama-3.1-8b, Zephyr-7b-beta, and Mistral-7b. Significant improvements were observed after fine-tuning, with notable enhancements in both cross-entropy and Dirichlet scores, particularly in the latter. Notably, after fine-tuning, the performance of Mistral-7B-Instruct-v0.3 was on par with GPT-4o. However, the experimental models that were examined were all still outperformed by the BERT and RoBERTa models in terms of cross-entropy scores. Our fine-tuning approach, which improves the model's understanding of text distributions in a text-to-text format, effectively enhances performance and alignment in complex decision-making contexts, underscoring the need for further research to refine ethical reasoning techniques and capture human judgment nuances.

cs.CL↗

arXiv2024

NusaMT-7B: Machine Translation for Low-Resource Indonesian Languages with Large Language Models

Large Language Models (LLMs) have demonstrated exceptional promise in translation tasks for high-resource languages. However, their performance in low-resource languages is limited by the scarcity of both parallel and monolingual corpora, as well as the presence of noise. Consequently, such LLMs suffer with alignment and have lagged behind State-of-The-Art (SoTA) neural machine translation (NMT) models in these settings. This paper introduces NusaMT-7B, an LLM-based machine translation model for low-resource Indonesian languages, starting with Balinese and Minangkabau. Leveraging the pretrained LLaMA2-7B, our approach integrates continued pre-training on monolingual data, Supervised Fine-Tuning (SFT), self-learning, and an LLM-based data cleaner to reduce noise in parallel sentences. In the FLORES-200 multilingual translation benchmark, NusaMT-7B outperforms SoTA models in the spBLEU metric by up to +6.69 spBLEU in translations into Balinese and Minangkabau, but underperforms by up to -3.38 spBLEU in translations into higher-resource languages. Our results show that fine-tuned LLMs can enhance translation quality for low-resource languages, aiding in linguistic preservation and cross-cultural communication.

cs.CL↗

arXiv2024

Time-series Crime Prediction Across the United States Based on Socioeconomic and Political Factors

Traditional crime prediction techniques are slow and inefficient when generating predictions as crime increases rapidly \cite{r15}. To enhance traditional crime prediction methods, a Long Short-Term Memory and Gated Recurrent Unit model was constructed using datasets involving gender ratios, high school graduation rates, political status, unemployment rates, and median income by state over multiple years. While there may be other crime prediction tools, personalizing the model with hand picked factors allows a unique gap for the project. Producing an effective model would allow policymakers to strategically allocate specific resources and legislation in geographic areas that are impacted by crime, contributing to the criminal justice field of research \cite{r2A}. The model has an average total loss value of 70.792.30, and a average percent error of 9.74 percent, however both of these values are impacted by extreme outliers and with the correct optimization may be corrected.

cs.LG↗

arXiv2024

Chain-of-Thought Augmentation with Logit Contrast for Enhanced Reasoning in Language Models

Rapidly increasing model scales coupled with steering methods such as chain-of-thought prompting have led to drastic improvements in language model reasoning. At the same time, models struggle with compositional generalization and are far from human performance on many reasoning-based benchmarks. Leveraging the success of chain-of-thought prompting, and also taking inspiration from context-aware decoding (CAD), we explore input-based contrasting methods to further encourage the type of reasoning induced by chain-of-thought prompting. While work remains to stabilize these results across datasets and models, the improvements we find warrant further investigation into input-based steering methods for context-aware reasoning.

cs.CL↗

arXiv2024

Enhancing Robustness of Human Detection Algorithms in Maritime SAR through Augmented Aerial Images to Simulate Weather Conditions

7,651 cases of Search and Rescue Missions (SAR) were reported by the United States Coast Guard in 2024, with over 1322 SAR helicopters deployed in the 6 first months alone. Through the utilizations of YOLO, we were able to run different weather conditions and lighting from our augmented dataset for training. YOLO then utilizes CNNs to apply a series of convolutions and pooling layers to the input image, where the convolution layers are able to extract the main features of the image. Through this, our YOLO model is able to learn to differentiate different objects which may considerably improve its accuracy, possibly enhancing the efficiency of SAR operations through enhanced detection accuracy. This paper aims to improve the model's accuracy of human detection in maritime SAR by evaluating a robust datasets containing various elevations and geological locations, as well as through data augmentation which simulates different weather and lighting. We observed that models trained on augmented datasets outperformed their non-augmented counterparts in which the human recall scores ranged from 0.891 to 0.911 with an improvement rate of 3.4\% on the YOLOv5l model. Results showed that these models demonstrate greater robustness to real-world conditions in varying of weather, brightness, tint, and contrast.

cs.CV↗

arXiv2024

Improving Water Quality Time-Series Prediction in Hong Kong using Sentinel-2 MSI Data and Google Earth Engine Cloud Computing

Effective water quality monitoring in coastal regions is crucial due to the progressive deterioration caused by pollution and human activities. To address this, this study develops time-series models to predict chlorophyll-a (Chl-a), suspended solids (SS), and turbidity using Sentinel-2 satellite data and Google Earth Engine (GEE) in the coastal regions of Hong Kong. Leveraging Long Short-Term Memory (LSTM) Recurrent Neural Networks, the study incorporates extensive temporal datasets to enhance prediction accuracy. The models utilize spectral data from Sentinel-2, focusing on optically active components, and demonstrate that selected variables closely align with the spectral characteristics of Chl-a and SS. The results indicate improved predictive performance over previous methods, highlighting the potential for remote sensing technology in continuous and comprehensive water quality assessment.

cs.LG↗

arXiv2024

Enhancing Depression Diagnosis with Chain-of-Thought Prompting

When using AI to detect signs of depressive disorder, AI models habitually draw preemptive conclusions. We theorize that using chain-of-thought (CoT) prompting to evaluate Patient Health Questionnaire-8 (PHQ-8) scores will improve the accuracy of the scores determined by AI models. In our findings, when the models reasoned with CoT, the estimated PHQ-8 scores were consistently closer on average to the accepted true scores reported by each participant compared to when not using CoT. Our goal is to expand upon AI models' understanding of the intricacies of human conversation, allowing them to more effectively assess a patient's feelings and tone, therefore being able to more accurately discern mental disorder symptoms; ultimately, we hope to augment AI models' abilities, so that they can be widely accessible and used in the medical field.

cs.CL↗

arXiv2024

From Bias to Balance: Detecting Facial Expression Recognition Biases in Large Multimodal Foundation Models

This study addresses the racial biases in facial expression recognition (FER) systems within Large Multimodal Foundation Models (LMFMs). Despite advances in deep learning and the availability of diverse datasets, FER systems often exhibit higher error rates for individuals with darker skin tones. Existing research predominantly focuses on traditional FER models (CNNs, RNNs, ViTs), leaving a gap in understanding racial biases in LMFMs. We benchmark four leading LMFMs: GPT-4o, PaliGemma, Gemini, and CLIP to assess their performance in facial emotion detection across different racial demographics. A linear classifier trained on CLIP embeddings obtains accuracies of 95.9\% for RADIATE, 90.3\% for Tarr, and 99.5\% for Chicago Face. Furthermore, we identify that Anger is misclassified as Disgust 2.1 times more often in Black Females than White Females. This study highlights the need for fairer FER systems and establishes a foundation for developing unbiased, accurate FER technologies. Visit https://kvjvhub.github.io/FERRacialBias/ for further information regarding the biases within facial expression recognition.

cs.CV↗

arXiv2024

Intraoperative Glioma Segmentation with YOLO + SAM for Improved Accuracy in Tumor Resection

Gliomas, a common type of malignant brain tumor, present significant surgical challenges due to their similarity to healthy tissue. Preoperative Magnetic Resonance Imaging (MRI) images are often ineffective during surgery due to factors such as brain shift, which alters the position of brain structures and tumors. This makes real-time intraoperative MRI (ioMRI) crucial, as it provides updated imaging that accounts for these shifts, ensuring more accurate tumor localization and safer resections. This paper presents a deep learning pipeline combining You Only Look Once Version 8 (YOLOv8) and Segment Anything Model Vision Transformer-base (SAM ViT-b) to enhance glioma detection and segmentation during ioMRI. Our model was trained using the Brain Tumor Segmentation 2021 (BraTS 2021) dataset, which includes standard magnetic resonance imaging (MRI) images, and noise-augmented MRI images that simulate ioMRI images. Noised MRI images are harder for a deep learning pipeline to segment, but they are more representative of surgical conditions. Achieving a Dice Similarity Coefficient (DICE) score of 0.79, our model performs comparably to state-of-the-art segmentation models tested on noiseless data. This performance demonstrates the model's potential to assist surgeons in maximizing tumor resection and improving surgical outcomes.

eess.IV↗

arXiv2024

Question-Analysis Prompting Improves LLM Performance in Reasoning Tasks

Although LLMs have the potential to transform many fields, they still underperform humans in reasoning tasks. Existing methods induce the model to produce step-by-step calculations, but this research explores the question: Does making the LLM analyze the question improve its performance? We propose a novel prompting strategy called Question Analysis Prompting (QAP), in which the model is prompted to explain the question in $n$ words before solving. The value of $n$ influences the length of response generated by the model. QAP is evaluated on GPT 3.5 Turbo and GPT 4 Turbo on arithmetic datasets GSM8K, AQuA, and SAT and commonsense dataset StrategyQA. QAP is compared with other state-of-the-art prompts including Chain-of-Thought (CoT), Plan and Solve Prompting (PS+) and Take A Deep Breath (TADB). QAP outperforms all state-of-the-art prompts on AQuA and SAT datasets on both GPT3.5 and GPT4. QAP consistently ranks among the top-2 prompts on 75\% of the tests. A key factor of QAP performance can be attributed to response length, where detailed responses are beneficial when answering harder questions, but can negatively affect easy questions.

cs.CL↗

arXiv2023

Lighter-Than-Air Autonomous Ball Capture and Scoring Robot -- Design, Development, and Deployment

This paper describes the full end-to-end design of our primary scoring agent in an aerial autonomous robotics competition from April 2023. As open-ended robotics competitions become more popular, we wish to begin documenting successful team designs and approaches. The intended audience of this paper is not only any future or potential participant in this particular national Defend The Republic (DTR) competition, but rather anyone thinking about designing their first robot or system to be entered in a competition with clear goals. Future DTR participants can and should either build on the ideas here, or find new alternate strategies that can defeat the most successful design last time. For non-DTR participants but students interested in robotics competitions, identifying the minimum viable system needed to be competitive is still important in helping manage time and prioritizing tasks that are crucial to competition success first.

cs.RO↗

arXiv2023

Atom-by-atom protein generation and beyond with language models

Protein language models learn powerful representations directly from sequences of amino acids. However, they are constrained to generate proteins with only the set of amino acids represented in their vocabulary. In contrast, chemical language models learn atom-level representations of smaller molecules that include every atom, bond, and ring. In this work, we show that chemical language models can learn atom-level representations of proteins enabling protein generation unconstrained to the standard genetic code and far beyond it. In doing so, we show that language models can generate entire proteins atom by atom -- effectively learning the multiple hierarchical layers of molecular information that define proteins from their primary sequence to their secondary, and tertiary structure. We demonstrate language models are able to explore beyond protein space -- generating proteins with modified sidechains that form unnatural amino acids. Even further, we find that language models can explore chemical space and protein space simultaneously and generate novel examples of protein-drug conjugates. The results demonstrate the potential for biomolecular design at the atom level using language models.

q-bio.BM↗

arXiv2023

Zespol: A Lightweight Environment for Training Swarming Agents

Agent-based modeling (ABM) and simulation have emerged as important tools for studying emergent behaviors, especially in the context of swarming algorithms for robotic systems. Despite significant research in this area, there is a lack of standardized simulation environments, which hinders the development and deployment of real-world robotic swarms. To address this issue, we present Zespol, a modular, Python-based simulation environment that enables the development and testing of multi-agent control algorithms. Zespol provides a flexible and extensible sandbox for initial research, with the potential for scaling to real-world applications. We provide a topological overview of the system and detailed descriptions of its plug-and-play elements. We demonstrate the fidelity of Zespol in simulated and real-word robotics by replicating existing works highlighting the simulation to real gap with the milling behavior. We plan to leverage Zespol's plug-and-play feature for neuromorphic computing in swarming scenarios, which involves using the modules in Zespol to simulate the behavior of neurons and their connections as synapses. This will enable optimizing and studying the emergent behavior of swarm systems in complex environments. Our goal is to gain a better understanding of the interplay between environmental factors and neural-like computations in swarming systems.

cs.RO↗

arXiv2023

Simulate Less, Expect More: Bringing Robot Swarms to Life via Low-Fidelity Simulations

This paper proposes a novel methodology for addressing the simulation-reality gap for multi-robot swarm systems. Rather than immediately try to shrink or `bridge the gap' anytime a real-world experiment failed that worked in simulation, we characterize conditions under which this is actually necessary. When these conditions are not satisfied, we show how very simple simulators can still be used to both (i) design new multi-robot systems, and (ii) guide real-world swarming experiments towards certain emergent behaviors when the gap is very large. The key ideas are an iterative simulator-in-the-design-loop in which real-world experiments, simulator modifications, and simulated experiments are intimately coupled in a way that minds the gap without needing to shrink it, as well as the use of minimally viable phase diagrams to guide real world experiments. We demonstrate the usefulness of our methods on deploying a real multi-robot swarm system to successfully exhibit an emergent milling behavior.

cs.RO↗

arXiv2022

FenceNet: Fine-grained Footwork Recognition in Fencing

Current data analysis for the Canadian Olympic fencing team is primarily done manually by coaches and analysts. Due to the highly repetitive, yet dynamic and subtle movements in fencing, manual data analysis can be inefficient and inaccurate. We propose FenceNet as a novel architecture to automate the classification of fine-grained footwork techniques in fencing. FenceNet takes 2D pose data as input and classifies actions using a skeleton-based action recognition approach that incorporates temporal convolutional networks to capture temporal information. We train and evaluate FenceNet on the Fencing Footwork Dataset (FFD), which contains 10 fencers performing 6 different footwork actions for 10-11 repetitions each (652 total videos). FenceNet achieves 85.4% accuracy under 10-fold cross-validation, where each fencer is left out as the test set. This accuracy is within 1% of the current state-of-the-art method, JLJA (86.3%), which selects and fuses features engineered from skeleton data, depth videos, and inertial measurement units. BiFenceNet, a variant of FenceNet that captures the "bidirectionality" of human movement through two separate networks, achieves 87.6% accuracy, outperforming JLJA. Since neither FenceNet nor BiFenceNet requires data from wearable sensors, unlike JLJA, they could be directly applied to most fencing videos, using 2D pose data as input extracted from off-the-shelf 2D human pose estimators. In comparison to JLJA, our methods are also simpler as they do not require manual feature engineering, selection, or fusion.

cs.CV↗

arXiv2021

Keeping it Simple: Language Models can learn Complex Molecular Distributions

Deep generative models of molecules have grown immensely in popularity, trained on relevant datasets, these models are used to search through chemical space. The downstream utility of generative models for the inverse design of novel functional compounds depends on their ability to learn a training distribution of molecules. The most simple example is a language model that takes the form of a recurrent neural network and generates molecules using a string representation. More sophisticated are graph generative models, which sequentially construct molecular graphs and typically achieve state of the art results. However, recent work has shown that language models are more capable than once thought, particularly in the low data regime. In this work, we investigate the capacity of simple language models to learn distributions of molecules. For this purpose, we introduce several challenging generative modeling tasks by compiling especially complex distributions of molecules. On each task, we evaluate the ability of language models as compared with two widely used graph generative models. The results demonstrate that language models are powerful generative models, capable of adeptly learning complex molecular distributions -- and yield better performance than the graph models. Language models can accurately generate: distributions of the highest scoring penalized LogP molecules in ZINC15, multi-modal molecular distributions as well as the largest molecules in PubChem.

cs.LG↗

arXiv2019

Understanding Childhood Vulnerability in The City of Surrey

Understanding the community conditions that best support universal access and improved childhood outcomes allows ultimately to improve decision-making in the areas of planning and investment across the early stages of childhood development. Here we describe two different data-driven approaches to visualizing the lived experiences of children throughout the City of Surrey, combining data derived from both public and private sources. In one approach, we find specifically that the Early Development Instrument measuring childhood vulnerabilities across varying domains can be used to cluster neighborhoods, and that census variables can help explain similarities between neighborhoods within these clusters. In our second approach, we use program registration data from the City of Surrey's Community and Recreation Services Division. We also find a critical age of entry and exit for each program related to early childhood development and beyond, and find that certain neighborhoods and recreational programs have larger retention rates than others. This report details the journey of using data to tell the story of these neighborhoods, and provides a lens to which community initiatives can be strategically crafted through their use.

cs.CY↗