arXiv subjects

Seunghyun Lee

Publications and source records attributed to Seunghyun Lee.

At least 37 records · Page 2Linked to original sources

arXiv2026

Parametric Mean-Field empirical Bayes in high-dimensional linear regression

In this paper, we consider the problem of parametric empirical Bayes estimation of an i.i.d. prior in high-dimensional Bayesian linear regression, with random design. We obtain the asymptotic distribution of the variational Empirical Bayes (vEB) estimator, which approximately maximizes a variational lower bound of the intractable marginal likelihood. We characterize a sharp phase transition behavior for the vEB estimator -- namely that it is information theoretically optimal (in terms of limiting variance) up to $p=o(n^{2/3})$ while it suffers from a sub-optimal convergence rate in higher dimensions. In the first regime, i.e., when $p=o(n^{2/3})$, we show how the estimated prior can be calibrated to enable valid coordinate-wise and delocalized inference, both under the \emph{empirical Bayes posterior} and the oracle posterior. In the second regime, we propose a debiasing technique as a way to improve the performance of the vEB estimator beyond $p=o(n^{2/3})$. Extensive numerical experiments corroborate our theoretical findings.

math.ST↗

arXiv2026

LocoVLM: Grounding Vision and Language for Adapting Versatile Legged Locomotion Policies

Recent advances in legged locomotion learning are still dominated by the utilization of geometric representations of the environment, limiting the robot's capability to respond to higher-level semantics such as human instructions. To address this limitation, we propose a novel approach that integrates high-level commonsense reasoning from foundation models into the process of legged locomotion adaptation. Specifically, our method utilizes a pre-trained large language model to synthesize an instruction-grounded skill database tailored for legged robots. A pre-trained vision-language model is employed to extract high-level environmental semantics and ground them within the skill database, enabling real-time skill advisories for the robot. To facilitate versatile skill control, we train a style-conditioned policy capable of generating diverse and robust locomotion skills with high fidelity to specified styles. To the best of our knowledge, this is the first work to demonstrate real-time adaptation of legged locomotion using high-level reasoning from environmental semantics and instructions with instruction-following accuracy of up to 87% without the need for online query to on-the-cloud foundation models.

cs.RO↗

arXiv2026

AI Twin: Enhancing ESL Speaking Practice through AI Self-Clones of a Better Me

Advances in AI have enabled ESL learners to practice speaking through conversational systems. However, most tools rely on explicit correction, which can interrupt the conversation and undermine confidence. Grounded in second language acquisition and motivational psychology, we present AI Twin, a system that rephrases learner utterances into more fluent English and delivers them in the learner's voice. Embodying a more confident and proficient version of the learner, AI Twin reinforces motivation through alignment with their aspirational Ideal L2 Self. Also, its use of implicit feedback through rephrasing preserves conversational flow and fosters an emotionally supportive environment. In a within-subject study with 20 adult ESL learners, we compared AI Twin with explicit correction and a non-personalized rephrasing agent. Results show that AI Twin elicited higher emotional engagement, with participants describing the experience as more motivating. These findings highlight the potential of self-representative AI for personalized, psychologically grounded support in ESL learning.

cs.HC↗

arXiv2026

On Theoretical Identifiability of Discrete Latent Causal Graphical Models

This paper considers a challenging problem of identifying a causal graphical model under the presence of latent variables. While various identifiability conditions have been proposed in the literature, they often require multiple pure children per latent variable or restrictions on the latent causal graph. Furthermore, it is common for all observed variables to exhibit the same modality. Consequently, the existing identifiability conditions are often too stringent for complex real-world data. We consider a general nonparametric measurement model with arbitrary observed variable types and binary latent variables, and propose a double triangular graphical condition that guarantees identifiability of the entire causal graphical model. The proposed condition significantly relaxes the popular pure children condition. We also establish necessary conditions for identifiability and provide valuable insights into fundamental limits of identifiability. Simulation studies verify that latent structures satisfying our conditions can be accurately estimated from data.

stat.ML↗

arXiv2026

Solar Open Technical Report

We introduce Solar Open, a 102B-parameter bilingual Mixture-of-Experts language model for underserved languages. Solar Open demonstrates a systematic methodology for building competitive LLMs by addressing three interconnected challenges. First, to train effectively despite data scarcity for underserved languages, we synthesize 4.5T tokens of high-quality, domain-specific, and RL-oriented data. Second, we coordinate this data through a progressive curriculum jointly optimizing composition, quality thresholds, and domain coverage across 20 trillion tokens. Third, to enable reasoning capabilities through scalable RL, we apply our proposed framework SnapPO for efficient optimization. Across benchmarks in English and Korean, Solar Open achieves competitive performance, demonstrating the effectiveness of this methodology for underserved language AI development.

cs.CL↗

arXiv2026

SPIO: Ensemble and Selective Strategies via LLM-Based Multi-Agent Planning in Automated Data Science

Large Language Models (LLMs) have enabled dynamic reasoning in automated data analytics, yet recent multi-agent systems remain limited by rigid, single-path workflows that restrict strategic exploration and often lead to suboptimal outcomes. To overcome these limitations, we propose SPIO (Sequential Plan Integration and Optimization), a framework that replaces rigid workflows with adaptive, multi-path planning across four core modules: data preprocessing, feature engineering, model selection, and hyperparameter tuning. In each module, specialized agents generate diverse candidate strategies, which are cascaded and refined by an optimization agent. SPIO offers two operating modes: SPIO-S for selecting a single optimal pipeline, and SPIO-E for ensembling top-k pipelines to maximize robustness. Extensive evaluations on Kaggle and OpenML benchmarks show that SPIO consistently outperforms state-of-the-art baselines, achieving an average performance gain of 5.6%. By explicitly exploring and integrating multiple solution paths, SPIO delivers a more flexible, accurate, and reliable foundation for automated data science.

cs.AI↗

arXiv2025

Question-to-Knowledge (Q2K): Multi-Agent Generation of Inspectable Facts for Product Mapping

Identifying whether two product listings refer to the same Stock Keeping Unit (SKU) is a persistent challenge in ecommerce, especially when explicit identifiers are missing and product names vary widely across platforms. Rule based heuristics and keyword similarity often misclassify products by overlooking subtle distinctions in brand, specification, or bundle configuration. To overcome these limitations, we propose Question to Knowledge (Q2K), a multi agent framework that leverages Large Language Models (LLMs) for reliable SKU mapping. Q2K integrates: (1) a Reasoning Agent that generates targeted disambiguation questions, (2) a Knowledge Agent that resolves them via focused web searches, and (3) a Deduplication Agent that reuses validated reasoning traces to reduce redundancy and ensure consistency. A human in the loop mechanism further refines uncertain cases. Experiments on real world consumer goods datasets show that Q2K surpasses strong baselines, achieving higher accuracy and robustness in difficult scenarios such as bundle identification and brand origin disambiguation. By reusing retrieved reasoning instead of issuing repeated searches, Q2K balances accuracy with efficiency, offering a scalable and interpretable solution for product integration.

cs.AI↗

arXiv2025

CLIP Meets Diffusion: A Synergistic Approach to Anomaly Detection

Anomaly detection is a complex problem due to the ambiguity in defining anomalies, the diversity of anomaly types (e.g., local and global defect), and the scarcity of training data. As such, it necessitates a comprehensive model capable of capturing both low-level and high-level features, even with limited data. To address this, we propose CLIPFUSION, a method that leverages both discriminative and generative foundation models. Specifically, the CLIP-based discriminative model excels at capturing global features, while the diffusion-based generative model effectively captures local details, creating a synergistic and complementary approach. Notably, we introduce a methodology for utilizing cross-attention maps and feature maps extracted from diffusion models specifically for anomaly detection. Experimental results on benchmark datasets (MVTec-AD, VisA) demonstrate that CLIPFUSION consistently outperforms baseline methods, achieving outstanding performance in both anomaly segmentation and classification. We believe that our method underscores the effectiveness of multi-modal and multi-model fusion in tackling the multifaceted challenges of anomaly detection, providing a scalable solution for real-world applications.

cs.CV↗

arXiv2025

A Study on the Framework for Evaluating the Ethics and Trustworthiness of Generative AI

This study provides an in_depth analysis of the ethical and trustworthiness challenges emerging alongside the rapid advancement of generative artificial intelligence (AI) technologies and proposes a comprehensive framework for their systematic evaluation. While generative AI, such as ChatGPT, demonstrates remarkable innovative potential, it simultaneously raises ethical and social concerns, including bias, harmfulness, copyright infringement, privacy violations, and hallucination. Current AI evaluation methodologies, which mainly focus on performance and accuracy, are insufficient to address these multifaceted issues. Thus, this study emphasizes the need for new human_centered criteria that also reflect social impact. To this end, it identifies key dimensions for evaluating the ethics and trustworthiness of generative AI_fairness, transparency, accountability, safety, privacy, accuracy, consistency, robustness, explainability, copyright and intellectual property protection, and source traceability and develops detailed indicators and assessment methodologies for each. Moreover, it provides a comparative analysis of AI ethics policies and guidelines in South Korea, the United States, the European Union, and China, deriving key approaches and implications from each. The proposed framework applies across the AI lifecycle and integrates technical assessments with multidisciplinary perspectives, thereby offering practical means to identify and manage ethical risks in real_world contexts. Ultimately, the study establishes an academic foundation for the responsible advancement of generative AI and delivers actionable insights for policymakers, developers, users, and other stakeholders, supporting the positive societal contributions of AI technologies.

cs.CY↗

arXiv2025

Inference on Gaussian mixture models with dependent labels

Gaussian mixture models are widely used to model data generated from multiple latent sources. Despite its popularity, most theoretical research assumes that the labels are either independent and identically distributed, or follows a Markov chain. It remains unclear how the fundamental limits of estimation change under more complex dependence. In this paper, we address this question for the spherical two-component Gaussian mixture model. We first show that for labels with an arbitrary dependence, a naive estimator based on the misspecified likelihood is $\sqrt{n}$-consistent. Additionally, under labels that follow an Ising model, we establish the information theoretic limitations for estimation, and discover an interesting phase transition as dependence becomes stronger. When the dependence is smaller than a threshold, the optimal estimator and its limiting variance exactly matches the independent case, for a wide class of Ising models. On the other hand, under stronger dependence, estimation becomes easier and the naive estimator is no longer optimal. Hence, we propose an alternative estimator based on the variational approximation of the likelihood, and argue its optimality under a specific Ising model.

math.ST↗

arXiv2025

Joint Learning of Pose Regression and Denoising Diffusion with Score Scaling Sampling for Category-level 6D Pose Estimation

Latest diffusion models have shown promising results in category-level 6D object pose estimation by modeling the conditional pose distribution with depth image input. The existing methods, however, suffer from slow convergence during training, learning its encoder with the diffusion denoising network in end-to-end fashion, and require an additional network that evaluates sampled pose hypotheses to filter out low-quality pose candidates. In this paper, we propose a novel pipeline that tackles these limitations by two key components. First, the proposed method pretrains the encoder with the direct pose regression head, and jointly learns the networks via the regression head and the denoising diffusion head, significantly accelerating training convergence while achieving higher accuracy. Second, sampling guidance via time-dependent score scaling is proposed s.t. the exploration-exploitation trade-off is effectively taken, eliminating the need for the additional evaluation network. The sampling guidance maintains multi-modal characteristics of symmetric objects at early denoising steps while ensuring high-quality pose generation at final steps. Extensive experiments on multiple benchmarks including REAL275, HouseCat6D, and ROPE, demonstrate that the proposed method, simple yet effective, achieves state-of-the-art accuracies even with single-pose inference, while being more efficient in both training and inference.

cs.CV↗

arXiv2025

Better by Comparison: Retrieval-Augmented Contrastive Reasoning for Automatic Prompt Optimization

Automatic prompt optimization has recently emerged as a strategy for improving the quality of prompts used in Large Language Models (LLMs), with the goal of generating more accurate and useful responses. However, most prior work focuses on direct prompt refinement or model fine-tuning, overlooking the potential of leveraging LLMs' inherent reasoning capability to learn from contrasting examples. In this paper, we present Contrastive Reasoning Prompt Optimization (CRPO), a novel framework that formulates prompt optimization as a retrieval-augmented reasoning process. Our approach retrieves top k reference prompt-response pairs from the HelpSteer2 dataset, an open source collection where each response is annotated for helpfulness, correctness, coherence, complexity, and verbosity, and constructs two complementary optimization paradigms: (1) tiered contrastive reasoning, where the LLM compares high-, medium-, and low-quality exemplars (both prompts and responses) to refine its own generation through reflective reasoning, and (2) multi-metric contrastive reasoning, where the LLM analyzes the best exemplars along each evaluation dimension and integrates their strengths into an optimized prompt. By explicitly contrasting high and low quality exemplars, CRPO enables the model to deduce why certain prompts succeed while others fail, thereby achieving more robust and interpretable optimization. Experimental results on the HelpSteer2 benchmark demonstrate that CRPO significantly outperforms baselines. Our findings highlight the promise of contrastive, retrieval-augmented reasoning for advancing automatic prompt optimization.

cs.CL↗

arXiv2025

Better Safe Than Sorry? Overreaction Problem of Vision Language Models in Visual Emergency Recognition

Vision-Language Models (VLMs) have shown capabilities in interpreting visual content, but their reliability in safety-critical scenarios remains insufficiently explored. We introduce VERI, a diagnostic benchmark comprising 200 synthetic images (100 contrastive pairs) and an additional 50 real-world images (25 pairs) for validation. Each emergency scene is paired with a visually similar but safe counterpart through human verification. Using a two-stage evaluation protocol (risk identification and emergency response), we assess 17 VLMs across medical emergencies, accidents, and natural disasters. Our analysis reveals an "overreaction problem": models achieve high recall (70-100%) but suffer from low precision, misclassifying 31-96% of safe situations as dangerous. Seven safe scenarios were universally misclassified by all models. This "better-safe-than-sorry" bias stems from contextual overinterpretation (88-98% of errors). Both synthetic and real-world datasets confirm these systematic patterns, challenging VLM reliability in safety-critical applications. Addressing this requires enhanced contextual reasoning in ambiguous visual situations.

cs.CV↗

arXiv2025

A New Query Expansion Approach via Agent-Mediated Dialogic Inquiry

Query expansion is widely used in Information Retrieval (IR) to improve search outcomes by supplementing initial queries with richer information. While recent Large Language Model (LLM) based methods generate pseudo-relevant content and expanded terms via multiple prompts, they often yield homogeneous, narrow expansions that lack the diverse context needed to retrieve relevant information. In this paper, we propose AMD: a new Agent-Mediated Dialogic Framework that engages in a dialogic inquiry involving three specialized roles: (1) a Socratic Questioning Agent reformulates the initial query into three sub-questions, with each question inspired by a specific Socratic questioning dimension, including clarification, assumption probing, and implication probing, (2) a Dialogic Answering Agent generates pseudo-answers, enriching the query representation with multiple perspectives aligned to the user's intent, and (3) a Reflective Feedback Agent evaluates and refines these pseudo-answers, ensuring that only the most relevant and informative content is retained. By leveraging a multi-agent process, AMD effectively crafts richer query representations through inquiry and feedback refinement. Extensive experiments on benchmarks including BEIR and TREC demonstrate that our framework outperforms previous methods, offering a robust solution for retrieval tasks.

cs.IR↗

arXiv2025

Deep Discrete Encoders: Identifiable Deep Generative Models for Rich Data with Discrete Latent Layers

In the era of generative AI, deep generative models (DGMs) with latent representations have gained tremendous popularity. Despite their impressive empirical performance, the statistical properties of these models remain underexplored. DGMs are often overparametrized, non-identifiable, and uninterpretable black boxes, raising serious concerns when deploying them in high-stakes applications. Motivated by this, we propose interpretable deep generative models for rich data types with discrete latent layers, called Deep Discrete Encoders (DDEs). A DDE is a directed graphical model with multiple binary latent layers. Theoretically, we propose transparent identifiability conditions for DDEs, which imply progressively smaller sizes of the latent layers as they go deeper. Identifiability ensures consistent parameter estimation and inspires an interpretable design of the deep architecture. Computationally, we propose a scalable estimation pipeline of a layerwise nonlinear spectral initialization followed by a penalized stochastic approximation EM algorithm. This procedure can efficiently estimate models with exponentially many latent components. Extensive simulation studies for high-dimensional data and deep architectures validate our theoretical results and demonstrate the excellent performance of our algorithms. We apply DDEs to three diverse real datasets with different data types to perform hierarchical topic modeling, image representation learning, and response time modeling in educational testing.

stat.ML↗

arXiv2025

AIAP: A No-Code Workflow Builder for Non-Experts with Natural Language and Multi-Agent Collaboration

While many tools are available for designing AI, non-experts still face challenges in clearly expressing their intent and managing system complexity. We introduce AIAP, a no-code platform that integrates natural language input with visual workflows. AIAP leverages a coordinated multi-agent system to decompose ambiguous user instructions into modular, actionable steps, hidden from users behind a unified interface. A user study involving 32 participants showed that AIAP's AI-generated suggestions, modular workflows, and automatic identification of data, actions, and context significantly improved participants' ability to develop services intuitively. These findings highlight that natural language-based visual programming significantly reduces barriers and enhances user experience in AI service design.

cs.HC↗

arXiv2025

CLT in high-dimensional Bayesian linear regression with low SNR

We study central limit theorems for linear statistics in high-dimensional Bayesian linear regression with product priors. Unlike the existing literature where the focus is on posterior contraction, we work under a non-contracting regime where neither the likelihood nor the prior dominates the other. This is motivated by modern high-dimensional datasets characterized by a bounded signal-to-noise ratio. This work takes a first step towards understanding limit distributions for one-dimensional projections of the posterior, as well as the posterior mean, in such regimes. Analogous to contractive settings, the resulting limiting distributions are Gaussian, but they heavily depend on the chosen prior and center around the Mean-Field approximation of the posterior. We study two concrete models of interest to illustrate this phenomenon -- the white noise design, and the (misspecified) Bayesian model. As an application, we construct credible intervals and compute their coverage probability under any misspecified prior. Our proofs rely on a combination of recent developments in Berry-Esseen type bounds for Random Field Ising models and both first and second order Poincaré inequalities. Notably, our results do not require any sparsity assumptions on the prior.

math.ST↗

arXiv2025

Inverse-Designed Metasurfaces for Wavefront Restoration in Under-Display Camera Systems

Under-display camera (UDC) systems enable full-screen displays in smartphones by embedding the camera beneath the display panel, eliminating the need for notches or punch holes. However, the periodic pixel structures of display panels introduce significant optical diffraction effects, leading to imaging artifacts and degraded visual quality. Conventional approaches to mitigate these distortions, such as deep learning-based image reconstruction, are often computationally expensive and unsuitable for real-time applications in consumer electronics. This work introduces an inverse-designed metasurface for wavefront restoration, addressing diffraction-induced distortions without relying on external software processing. The proposed metasurface effectively suppresses higher-order diffraction modes caused by the metallic pixel structures, restores the optical wavefront, and enhances imaging quality across multiple wavelengths. By eliminating the need for software-based post-processing, our approach establishes a scalable, real-time optical solution for diffraction management in UDC systems. This advancement paves the way to achieve software-free real-time image restoration frameworks for many industrial applications.

physics.optics↗