Jaewook Lee

Hunter McNichols, Wanyong Feng, Jaewook Lee et al.

Multiple-choice questions (MCQs) are ubiquitous in almost all levels of education since they are easy to administer, grade, and are a reliable form of assessment. An important aspect of MCQs is the distractors, i.e., incorrect options that are designed to target specific misconceptions or insufficient knowledge among students. To date, the task of crafting high-quality distractors has largely remained a labor-intensive process for teachers and learning content designers, which has limited scalability. In this work, we explore the task of automated distractor and corresponding feedback message generation in math MCQs using large language models. We establish a formulation of these two tasks and propose a simple, in-context learning-based solution. Moreover, we propose generative AI-based metrics for evaluating the quality of the feedback messages. We conduct extensive experiments on these tasks using a real-world MCQ dataset. Our findings suggest that there is a lot of room for improvement in automated distractor and feedback generation; based on these findings, we outline several directions for future work.

Hoki Kim, Jinseong Park, Yujin Choi et al.

Sharpness-aware minimization (SAM) is a recently proposed training method that seeks to find flat minima in deep learning, resulting in state-of-the-art performance across various domains. Instead of minimizing the loss of the current weights, SAM minimizes the worst-case loss in its neighborhood in the parameter space. In this paper, we demonstrate that SAM dynamics can have convergence instability that occurs near a saddle point. Utilizing the qualitative theory of dynamical systems, we explain how SAM becomes stuck in the saddle point and then theoretically prove that the saddle point can become an attractor under SAM dynamics. Additionally, we show that this convergence instability can also occur in stochastic dynamical systems by establishing the diffusion of SAM. We prove that SAM diffusion is worse than that of vanilla gradient descent in terms of saddle point escape. Further, we demonstrate that often overlooked training tricks, momentum and batch-size, are important to mitigate the convergence instability and achieve high generalization performance. Our theoretical and empirical results are thoroughly verified through experiments on several well-known optimization problems and benchmark tasks.

Jinseong Park, Seungyun Lee, Woojin Jeong et al.

Time series generation is widely used in real-world applications such as simulation, data augmentation, and hypothesis testing. Recently, diffusion models have emerged as the de facto approach to time series generation, enabling diverse synthesis scenarios. However, the fixed standard-Gaussian diffusion prior may be ill-suited for time series data, which exhibit properties such as temporal order and fixed time points. In this paper, we propose TimeBridge, a framework that flexibly synthesizes time series data by using diffusion bridges to learn paths between a chosen prior and the data distribution. We then explore several prior designs tailored to time series synthesis. Our framework covers (i) data- and time-dependent priors for unconditional generation and (ii) scale-preserving priors for conditional generation. Experiments show that our framework with data-driven priors outperforms standard diffusion models on time series generation.

Alexander Scarlatos, Naiming Liu, Jaewook Lee et al.

Generative artificial intelligence (AI) has the potential to scale up personalized tutoring through large language models (LLMs). Recent AI tutors are adapted for the tutoring task by training or prompting LLMs to follow effective pedagogical principles, though they are not trained to maximize student learning throughout the course of a dialogue. Therefore, they may engage with students in a suboptimal way. We address this limitation by introducing an approach to train LLMs to generate tutor utterances that maximize the likelihood of student correctness, while still encouraging the model to follow good pedagogical practice. Specifically, we generate a set of candidate tutor utterances and score them using (1) an LLM-based student model to predict the chance of correct student responses and (2) a pedagogical rubric evaluated by GPT-4o. We then use the resulting data to train an open-source LLM, Llama 3.1 8B, using direct preference optimization. We show that tutor utterances generated by our model lead to significantly higher chances of correct student responses while maintaining the pedagogical quality of GPT-4o. We also conduct qualitative analyses and a human evaluation to demonstrate that our model generates high quality tutor utterances.

Hoki Kim, Jinseong Park, Jaewook Lee

Adversarial attacks have verified the existence of the vulnerability of neural networks. By adding small perturbations to a benign example, adversarial attacks successfully generate adversarial examples that lead misclassification of deep learning models. More importantly, an adversarial example generated from a specific model can also deceive other models without modification. We call this phenomenon ``transferability". Here, we analyze the relationship between transferability and input transformation with additive noise by mathematically proving that the modified optimization can produce more transferable adversarial examples.

Hunter McNichols, Jaewook Lee, Stephen Fancsali et al.

Intelligent Tutoring Systems (ITSs) often contain an automated feedback component, which provides a predefined feedback message to students when they detect a predefined error. To such a feedback component, we often resort to template-based approaches. These approaches require significant effort from human experts to detect a limited number of possible student errors and provide corresponding feedback. This limitation is exemplified in open-ended math questions, where there can be a large number of different incorrect errors. In our work, we examine the capabilities of large language models (LLMs) to generate feedback for open-ended math questions, similar to that of an established ITS that uses a template-based approach. We fine-tune both open-source and proprietary LLMs on real student responses and corresponding ITS-provided feedback. We measure the quality of the generated feedback using text similarity metrics. We find that open-source and proprietary models both show promise in replicating the feedback they see during training, but do not generalize well to previously unseen student errors. These results suggest that despite being able to learn the formatting of feedback, LLMs are not able to fully understand mathematical errors made by students.

Donghoon Shin, Alice Gao, Rock Yuren Pang et al.

Generative AI is known for its tendency to homogenize, often reproducing dominant style conventions found in training data. However, it remains unclear how these homogenizing effects extend to complex structural tasks like web design. As lay creators increasingly turn to LLMs to 'vibe-code' websites -- prompting for aesthetic and functional goals rather than writing code -- they may inadvertently narrow the diversity of their designs, and limit creative expression throughout the internet. In this paper, we interrogate the possibility of design homogenization in web vibe coding. We first characterize the vibe coding lifecycle, pinpointing stages where homogenization risks may arise. We then conduct a sociotechnical risk analysis unpacking the potential harms of web vibe coding and their interaction with design homogenization. We identify that the push for frictionless generation can exacerbate homogenization and its harms. Finally, we propose a mitigation framework centered on the idea of productive friction. Through case studies at the micro, meso, and macro levels, we show how centering productive friction can empower creators to challenge default outputs and preserve diverse expression in AI-mediated web design.

Jaewook Lee, Digory Smith, Simon Woodhead et al.

Multiple choice questions (MCQs) are a popular method for evaluating students' knowledge due to their efficiency in administration and grading. Crafting high-quality math MCQs is a labor-intensive process that requires educators to formulate precise stems and plausible distractors. Recent advances in large language models (LLMs) have sparked interest in automating MCQ creation, but challenges persist in ensuring mathematical accuracy and addressing student errors. This paper introduces a prototype tool designed to facilitate collaboration between LLMs and educators for streamlining the math MCQ generation process. We conduct a pilot study involving math educators to investigate how the tool can help them simplify the process of crafting high-quality math MCQs. We found that while LLMs can generate well-formulated question stems, their ability to generate distractors that capture common student errors and misconceptions is limited. Nevertheless, a human-AI collaboration has the potential to enhance the efficiency and effectiveness of MCQ generation.

Yujin Choi, Youngjoo Park, Junyoung Byun et al.

Retrieval-augmented generation (RAG) mitigates the hallucination problem in large language models (LLMs) and has proven effective for personalized usages. However, delivering private retrieved documents directly to LLMs introduces vulnerability to membership inference attacks (MIAs), which try to determine whether the target data point exists in the private external database or not. Based on the insight that MIA queries typically exhibit high similarity to only one target document, we introduce a novel similarity-based MIA detection framework designed for the RAG system. With the proposed method, we show that a simple detect-and-hide strategy can successfully obfuscate attackers, maintain data utility, and remain system-agnostic against MIA. We experimentally prove its detection and defense against various state-of-the-art MIA methods and its adaptability to existing RAG systems.

Hoki Kim, Jinseong Park, Yujin Choi et al.

Neural additive model (NAM) is a recently proposed explainable artificial intelligence (XAI) method that utilizes neural network-based architectures. Given the advantages of neural networks, NAMs provide intuitive explanations for their predictions with high model performance. In this paper, we analyze a critical yet overlooked phenomenon: NAMs often produce inconsistent explanations, even when using the same architecture and dataset. Traditionally, such inconsistencies have been viewed as issues to be resolved. However, we argue instead that these inconsistencies can provide valuable explanations within the given data model. Through a simple theoretical framework, we demonstrate that these inconsistencies are not mere artifacts but emerge naturally in datasets with multiple important features. To effectively leverage this information, we introduce a novel framework, Bayesian Neural Additive Model (BayesNAM), which integrates Bayesian neural networks and feature dropout, with theoretical proof demonstrating that feature dropout effectively captures model inconsistencies. Our experiments demonstrate that BayesNAM effectively reveals potential problems such as insufficient data or structural limitations of the model, providing more reliable explanations and potential remedies.

Jaewook Lee, Xinyang Sun, Ethan Errington et al.

Accurate prediction of Global Warming Potential (GWP) is essential for assessing the environmental impact of chemical processes and materials. Traditional GWP prediction models rely predominantly on molecular structure, overlooking critical process-related information. In this study, we present an integrative GWP prediction model that combines molecular descriptors (MACCS keys and Mordred descriptors) with process information (process title, description, and location) to improve predictive accuracy and interpretability. Using a deep neural network (DNN) model, we achieved an R-squared of 86% on test data with Mordred descriptors, process location, and description information, representing a 25% improvement over the previous benchmark of 61%; XAI analysis further highlighted the significant role of process title embeddings in enhancing model predictions. To enhance interpretability, we employed a Kolmogorov-Arnold Network (KAN) to derive a symbolic formula for GWP prediction, capturing key molecular and process features and providing a transparent, interpretable alternative to black-box models, enabling users to gain insights into the molecular and process factors influencing GWP. Error analysis showed that the model performs reliably in densely populated data ranges, with increased uncertainty for higher GWP values. This analysis allows users to manage prediction uncertainty effectively, supporting data-driven decision-making in chemical and process design. Our results suggest that integrating both molecular and process-level information in GWP prediction models yields substantial gains in accuracy and interpretability, offering a valuable tool for sustainability assessments. Future work may extend this approach to additional environmental impact categories and refine the model to further enhance its predictive reliability.

Jaewook Lee, Andrew Lan

In second language vocabulary learning, existing works have primarily focused on either the learning interface or scheduling personalized retrieval practices to maximize memory retention. However, the learning content, i.e., the information presented on flashcards, has mostly remained constant. Keyword mnemonic is a notable learning strategy that relates new vocabulary to existing knowledge by building an acoustic and imagery link using a keyword that sounds alike. Beyond that, producing verbal and visual cues associated with the keyword to facilitate building these links requires a manual process and is not scalable. In this paper, we explore an opportunity to use large language models to automatically generate verbal and visual cues for keyword mnemonics. Our approach, an end-to-end pipeline for auto-generating verbal and visual cues, can automatically generate highly memorable cues. We investigate the effectiveness of our approach via a human participant experiment by comparing it with manually generated cues.

Hoki Kim, Woojin Lee, Sungyoon Lee et al.

Adversarial robustness is considered as a required property of deep neural networks. In this study, we discover that adversarially trained models might have significantly different characteristics in terms of margin and smoothness, even they show similar robustness. Inspired by the observation, we investigate the effect of different regularizers and discover the negative effect of the smoothness regularizer on maximizing the margin. Based on the analyses, we propose a new method called bridged adversarial training that mitigates the negative effect by bridging the gap between clean and adversarial examples. We provide theoretical and empirical evidence that the proposed method provides stable and better robustness, especially for large perturbations.

Sungyoon Lee, Hoki Kim, Jaewook Lee

Deep learning is vulnerable to adversarial examples. Many defenses based on randomized neural networks have been proposed to solve the problem, but fail to achieve robustness against attacks using proxy gradients such as the Expectation over Transformation (EOT) attack. We investigate the effect of the adversarial attacks using proxy gradients on randomized neural networks and demonstrate that it highly relies on the directional distribution of the loss gradients of the randomized neural network. We show in particular that proxy gradients are less effective when the gradients are more scattered. To this end, we propose Gradient Diversity (GradDiv) regularizations that minimize the concentration of the gradients to build a robust randomized neural network. Our experiments on MNIST, CIFAR10, and STL10 show that our proposed GradDiv regularizations improve the adversarial robustness of randomized neural networks against a variety of state-of-the-art attack methods. Moreover, our method efficiently reduces the transferability among sample models of randomized neural networks.

Hoki Kim, Woojin Lee, Jaewook Lee

Although fast adversarial training has demonstrated both robustness and efficiency, the problem of "catastrophic overfitting" has been observed. This is a phenomenon in which, during single-step adversarial training, the robust accuracy against projected gradient descent (PGD) suddenly decreases to 0% after a few epochs, whereas the robust accuracy against fast gradient sign method (FGSM) increases to 100%. In this paper, we demonstrate that catastrophic overfitting is very closely related to the characteristic of single-step adversarial training which uses only adversarial examples with the maximum perturbation, and not all adversarial examples in the adversarial direction, which leads to decision boundary distortion and a highly curved loss surface. Based on this observation, we propose a simple method that not only prevents catastrophic overfitting, but also overrides the belief that it is difficult to prevent multi-step adversarial attacks with single-step adversarial training.