arXiv subjects

Mingyu Lee

Publications and source records attributed to Mingyu Lee.

13 recordsLinked to original sources

arXiv2026

Uncertainty-Aware Structure-Property Mapping of Spinodoid Metamaterials via Heteroscedastic Gaussian Process Regression

Spinodoid metamaterials offer a broad, tunable design space for anisotropic mechanical properties, yet their structure-property relationships are commonly treated as representative mappings from cone-angle descriptors to single effective stiffness values. This deterministic view overlooks the stochastic nature of Gaussian random field (GRF)-based topology generation, where identical cone-angle descriptors can produce different morphology realizations and property scatter. Here, we present an uncertainty-aware structure-property mapping framework that reinterprets cone-angle descriptors as stochastic descriptors associated with input-dependent property distributions. Using heteroscedastic Gaussian process regression (GPR), the framework infers input-dependent predictive uncertainty from sparse one-realization-per-point data without requiring empirical variance labels at every design point. The results show that stiffness scatter differs across tensor components according to each component's mechanically active directions, and that parameter sets yielding identical mean stiffness can carry different aleatoric uncertainty. Applying this uncertainty to reliability-based design optimization (RBDO), we show that a deterministic optimum is highly susceptible to constraint violation once morphology-induced variability is considered, and that a homoscedastic RBDO formulation fails to meet the prescribed reliability target - only the heteroscedastic formulation satisfies the reliability target under the heteroscedastic uncertainty evaluation. This establishes uncertainty-aware surrogate modeling as essential for reliability-aware inverse design of spinodoid metamaterials; extending the framework to nonlinear responses remains for future work.

physics.comp-ph↗

arXiv2026

Polestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs

The inference efficiency of diffusion large language models (dLLMs) is constrained by two challenges: bidirectional attention precludes efficient KV-cache reuse, while increasing decoding parallelism with static confidence thresholds can compromise generation quality. We observe that both challenges arise from a shared phenomenon: as tokens are decoded, their contextual integration through bidirectional attention causes token representations to drift (evolve) across decoding steps. This insight motivates Polestar, a training-free inference framework that uses token representation drift as a unified signal to jointly address both challenges. Polestar comprises two components: Polestar-Cache, which identifies stale KV-cache positions via drift and performs sparse KV-cache refreshes to enable efficient reuse, and Polestar-Commit, which detects sharp drift events to reliably identify commit-ready tokens. Across mathematics and coding benchmarks on several dLLM families, Polestar sets a new state of the art on the accuracy-throughput Pareto frontier, achieving up to 10.73% accuracy improvement, up to 3.7x higher throughput, and high decoding parallelism of 3.67 tokens per forward pass over existing baselines.

cs.CL↗

arXiv2026

Fundamental Limitations of Post-Quantum Cryptographic Architectures

Modern lattice-based cryptography, particularly the learning with errors paradigm, relies on injecting artificial noise to secure data against quantum adversaries. This study systematically examines the theoretical and physical boundaries of this noise-reliant model across four interconnected domains: computational complexity, information-theoretic thermodynamics, quantum error correction, and quantum learning theory. Starting from the algorithmic foundation, our analysis notes that these frameworks rely on provisional complexity-theoretic assumptions that remain vulnerable to future quantum algorithmic advancements. Furthermore, by translating this cryptographic mechanism into physical thermodynamics, we illustrate that intentionally injected discrete Gaussian noise does not equate to the permanent erasure of information. Because the structural integrity of the cryptographic secret remains preserved within the ciphertext, advanced quantum error correction protocols and quantum learning models can efficiently extract the underlying mathematical kernel. Ultimately, we suggest that while lattice-based cryptography provides a robust transitional alternative, definitively classifying these frameworks as unconditionally post-quantum represents a premature classification relying on transient physical bottlenecks rather than impenetrable theoretical boundaries.

quant-ph↗

arXiv2025

OuroMamba: A Data-Free Quantization Framework for Vision Mamba

We present OuroMamba, the first data-free post-training quantization (DFQ) method for vision Mamba-based models (VMMs). We identify two key challenges in enabling DFQ for VMMs, (1) VMM's recurrent state transitions restricts capturing of long-range interactions and leads to semantically weak synthetic data, (2) VMM activations exhibit dynamic outlier variations across time-steps, rendering existing static PTQ techniques ineffective. To address these challenges, OuroMamba presents a two-stage framework: (1) OuroMamba-Gen to generate semantically rich and meaningful synthetic data. It applies contrastive learning on patch level VMM features generated through neighborhood interactions in the latent state space, (2) OuroMamba-Quant to employ mixed-precision quantization with lightweight dynamic outlier detection during inference. In specific, we present a thresholding based outlier channel selection strategy for activations that gets updated every time-step. Extensive experiments across vision and generative tasks show that our data-free OuroMamba surpasses existing data-driven PTQ techniques, achieving state-of-the-art performance across diverse quantization settings. Additionally, we implement efficient GPU kernels to achieve practical latency speedup of up to 2.36x. Code and synthetic dataset are available here: https://github.com/georgia-tech-synergy-lab/ICCV-OuroMamba

cs.CV↗

arXiv2024

MolTRES: Improving Chemical Language Representation Learning for Molecular Property Prediction

Chemical representation learning has gained increasing interest due to the limited availability of supervised data in fields such as drug and materials design. This interest particularly extends to chemical language representation learning, which involves pre-training Transformers on SMILES sequences -- textual descriptors of molecules. Despite its success in molecular property prediction, current practices often lead to overfitting and limited scalability due to early convergence. In this paper, we introduce a novel chemical language representation learning framework, called MolTRES, to address these issues. MolTRES incorporates generator-discriminator training, allowing the model to learn from more challenging examples that require structural understanding. In addition, we enrich molecular representations by transferring knowledge from scientific literature by integrating external materials embedding. Experimental results show that our model outperforms existing state-of-the-art models on popular molecular property prediction tasks.

physics.chem-ph↗

arXiv2024

Sub-symmetry Protected Topology in Topological Insulators and Superconductors

Exploration of topology protected by a certain symmetry is central in condensed matter physics. A recent idea of sub-symmetry-protected (SSP) topology--remains of a broken symmetry can still protect specific topological boundary states--has been developed and demonstrated in an optical system [Nat. Phys. 19, 992-998 (2023)]. Here, we extend this idea further by applying sub-symmetry-protecting perturbation (SSPP) to one-dimensional topological insulating and superconducting systems using the Su-Schrieffer-Hegger (SSH) and Kitaev models. Using the tight-binding and low-energy effective theory, we show that the SSP boundary states retain topological properties while the SSPP results in the asymmetry of boundary states. For the SSH model, an SSP zero-energy edge state localized on one edge possesses quantized polarization. In contrast, the other edge state is perturbed to have non-zero energy, and its polarization is not quantized. For topological superconductors, zero-energy SSP Majorana boundary states for spinful Kitaev models emerge on only one edge, contrary to the conventional belief that Majorana fermions emerge at opposite edges. Our findings can be used as a platform to expand our understanding of topological materials as they broaden our understanding of the symmetry in a topological system and a method to engineer Majorana fermions.

cond-mat.mes-hall↗

arXiv2024

A Highly Efficient Hybrid Fiber Optic Laser Using a Cesium Atom Vapor Cell as an Optical Gain Medium

A new scheme of a highly efficient hybrid laser cavity is proposed and experimentally demonstrated utilizing a hot cesium (Cs) vapor cell as an optical gain medium. The laser cavity consists of a macroscopic concave reflecting mirror (>99% reflectivity) and a 4% Fresnel-reflecting perpendicularly cleaved facet of a single mode fiber (SMF). The cylindrical cesium gain cell is located between these two reflectors. The SMF serves multiple roles: 1) a passive mode-matching component to approximate the pump beam diameter to that of the laser cavity mode within the cesium cell, 2) an output coupler with low reflectivity, and 3) a high beam-quality laser delivery with a low loss. Optimizing the pump beam waist diameter and the cesium vapor cell temperature, a high slope efficiency of 86% and continuous wave power of 419 mW were obtained in the pump power range of 400 to 600 mW, with an optical-to-optical conversion efficiency of 71%. The unique multi-functional role of the SMF in the hybrid cavity is fully described, and it can also be applied to other phases of high optical gain media.

physics.optics↗

arXiv2024

Text-Guided Variational Image Generation for Industrial Anomaly Detection and Segmentation

We propose a text-guided variational image generation method to address the challenge of getting clean data for anomaly detection in industrial manufacturing. Our method utilizes text information about the target object, learned from extensive text library documents, to generate non-defective data images resembling the input image. The proposed framework ensures that the generated non-defective images align with anticipated distributions derived from textual and image-based knowledge, ensuring stability and generality. Experimental results demonstrate the effectiveness of our approach, surpassing previous methods even with limited non-defective data. Our approach is validated through generalization tests across four baseline models and three distinct datasets. We present an additional analysis to enhance the effectiveness of anomaly detection models by utilizing the generated images.

cs.CV↗

arXiv2024

Disentangling quantum neural networks for unified estimation of quantum entropies and distance measures

The estimation of quantum entropies and distance measures, such as von Neumann entropy, R\'{e}nyi entropy, Tsallis entropy, trace distance, and fidelity-induced distances such as the Bures distance, has been a key area of research in quantum information science. In our study, we introduce the disentangling quantum neural network (DEQNN), designed to efficiently estimate various physical quantities in quantum information. Estimation algorithms for these quantities are generally tied to the size of the Hilbert space of the quantum state to be estimated. Our proposed DEQNN offers a unified dimensionality reduction methodology that can significantly reduce the size of the Hilbert space while preserving the values of diverse physical quantities. We provide an in-depth discussion of the physical scenarios and limitations in which our algorithm is applicable, as well as the learnability of the proposed quantum neural network.

quant-ph↗

arXiv2023

Improving Bias Mitigation through Bias Experts in Natural Language Understanding

Biases in the dataset often enable the model to achieve high performance on in-distribution data, while poorly performing on out-of-distribution data. To mitigate the detrimental effect of the bias on the networks, previous works have proposed debiasing methods that down-weight the biased examples identified by an auxiliary model, which is trained with explicit bias labels. However, finding a type of bias in datasets is a costly process. Therefore, recent studies have attempted to make the auxiliary model biased without the guidance (or annotation) of bias labels, by constraining the model's training environment or the capability of the model itself. Despite the promising debiasing results of recent works, the multi-class learning objective, which has been naively used to train the auxiliary model, may harm the bias mitigation effect due to its regularization effect and competitive nature across classes. As an alternative, we propose a new debiasing framework that introduces binary classifiers between the auxiliary model and the main model, coined bias experts. Specifically, each bias expert is trained on a binary classification task derived from the multi-class classification task via the One-vs-Rest approach. Experimental results demonstrate that our proposed strategy improves the bias identification ability of the auxiliary model. Consequently, our debiased model consistently outperforms the state-of-the-art on various challenge datasets.

cs.CL↗

arXiv2023

Mutual information maximizing quantum generative adversarial networks

One of the most promising applications in the era of Noisy Intermediate-Scale Quantum (NISQ) computing is quantum generative adversarial networks (QGANs), which offer significant quantum advantages over classical machine learning in various domains. However, QGANs suffer from mode collapse and lack explicit control over the features of generated outputs. To overcome these limitations, we propose InfoQGAN, a novel quantum-classical hybrid generative adversarial network that integrates the principles of InfoGAN with a QGAN architecture. Our approach employs a variational quantum circuit for data generation, a classical discriminator, and a Mutual Information Neural Estimator (MINE) to explicitly optimize the mutual information between latent codes and generated samples. Numerical simulations on synthetic 2D distributions and Iris dataset augmentation demonstrate that InfoQGAN effectively mitigates mode collapse while achieving robust feature disentanglement in the quantum generator. By leveraging these advantages, InfoQGAN not only enhances training stability but also improves data augmentation performance through controlled feature generation. These results highlight the potential of InfoQGAN as a foundational approach for advancing quantum generative modeling in the NISQ era.

quant-ph↗

arXiv2022

Efficient Pre-training of Masked Language Model via Concept-based Curriculum Masking

Masked language modeling (MLM) has been widely used for pre-training effective bidirectional representations, but incurs substantial training costs. In this paper, we propose a novel concept-based curriculum masking (CCM) method to efficiently pre-train a language model. CCM has two key differences from existing curriculum learning approaches to effectively reflect the nature of MLM. First, we introduce a carefully-designed linguistic difficulty criterion that evaluates the MLM difficulty of each token. Second, we construct a curriculum that gradually masks words related to the previously masked words by retrieving a knowledge graph. Experimental results show that CCM significantly improves pre-training efficiency. Specifically, the model trained with CCM shows comparative performance with the original BERT on the General Language Understanding Evaluation benchmark at half of the training cost.

cs.CL↗

arXiv2020

Deep Learning-Based Solvability of Underdetermined Inverse Problems in Medical Imaging

Recently, with the significant developments in deep learning techniques, solving underdetermined inverse problems has become one of the major concerns in the medical imaging domain. Typical examples include undersampled magnetic resonance imaging, interior tomography, and sparse-view computed tomography, where deep learning techniques have achieved excellent performances. Although deep learning methods appear to overcome the limitations of existing mathematical methods when handling various underdetermined problems, there is a lack of rigorous mathematical foundations that would allow us to elucidate the reasons for the remarkable performance of deep learning methods. This study focuses on learning the causal relationship regarding the structure of the training data suitable for deep learning, to solve highly underdetermined inverse problems. We observe that a majority of the problems of solving underdetermined linear systems in medical imaging are highly non-linear. Furthermore, we analyze if a desired reconstruction map can be learnable from the training data and underdetermined system.

eess.IV↗