arXiv subjects

Geonhee Han

Publications and source records attributed to Geonhee Han.

8 recordsLinked to original sources

arXiv2026

UNR-Explainer: Counterfactual Explanations for Unsupervised Node Representation Learning Models

Node representation learning, such as Graph Neural Networks (GNNs), has emerged as a pivotal method in machine learning. The demand for reliable explanation generation surges, yet unsupervised models remain underexplored. To bridge this gap, we introduce a method for generating counterfactual (CF) explanations in unsupervised node representation learning. We identify the most important subgraphs that cause a significant change in the k-nearest neighbors of a node of interest in the learned embedding space upon perturbation. The k-nearest neighbor-based CF explanation method provides simple, yet pivotal, information for understanding unsupervised downstream tasks, such as top-k link prediction and clustering. Consequently, we introduce UNR-Explainer for generating expressive CF explanations for Unsupervised Node Representation learning methods based on a Monte Carlo Tree Search (MCTS). The proposed method demonstrates superior performance on diverse datasets for unsupervised GraphSAGE and DGI.

cs.LG↗

arXiv2025

Image Diffusion Models Exhibit Emergent Temporal Propagation in Videos

Image diffusion models, though originally developed for image generation, implicitly capture rich semantic structures that enable various recognition and localization tasks beyond synthesis. In this work, we investigate their self-attention maps can be reinterpreted as semantic label propagation kernels, providing robust pixel-level correspondences between relevant image regions. Extending this mechanism across frames yields a temporal propagation kernel that enables zero-shot object tracking via segmentation in videos. We further demonstrate the effectiveness of test-time optimization strategies-DDIM inversion, textual inversion, and adaptive head weighting-in adapting diffusion features for robust and consistent label propagation. Building on these findings, we introduce DRIFT, a framework for object tracking in videos leveraging a pretrained image diffusion model with SAM-guided mask refinement, achieving state-of-the-art zero-shot performance on standard video object segmentation benchmarks.

cs.CV↗

arXiv2025

Efficient scenario analysis in real-time Bayesian election forecasting via sequential meta-posterior sampling

Bayesian aggregation lets election forecasters combine diverse sources of information, such as state polls and economic and political indicators: as in our collaboration with The Economist magazine. However, the demands of real-time posterior updating, model checking, and communication introduce practical methodological challenges. In particular, sensitivity and scenario analysis help trace intricate dependencies and understand model behavior. Yet, under standard Markov chain Monte Carlo, even small tweaks to the model (e.g., in priors, data, hyperparameters) require full refitting, making it computationally expensive in real time. We introduce a meta-modeling strategy paired with a sequential sampling scheme; by traversing posterior meta-models, we enable real-time structured analyses without repeated refitting. In a back-test of the model, we demonstrate substantial computational gains, and show how seemingly innocuous data-wrangling choices can unintentionally introduce partisan asymmetries into the model's mechanisms for handling systematic polling error. Code is available at https://github.com/geonhee619/SMC-Sense.

stat.ME↗

arXiv2025

DGMO: Training-Free Audio Source Separation through Diffusion-Guided Mask Optimization

Language-queried Audio Source Separation (LASS) enables open-vocabulary sound separation via natural language queries. While existing methods rely on task-specific training, we explore whether pretrained diffusion models, originally designed for audio generation, can inherently perform separation without further training. In this study, we introduce a training-free framework leveraging generative priors for zero-shot LASS. Analyzing naive adaptations, we identify key limitations arising from modality-specific challenges. To address these issues, we propose Diffusion-Guided Mask Optimization (DGMO), a test-time optimization framework that refines spectrogram masks for precise, input-aligned separation. Our approach effectively repurposes pretrained diffusion models for source separation, achieving competitive performance without task-specific supervision. This work expands the application of diffusion models beyond generation, establishing a new paradigm for zero-shot audio separation. The code is available at: https://wltschmrz.github.io/DGMO/

eess.AS↗

arXiv2025

Random Conditioning with Distillation for Data-Efficient Diffusion Model Compression

Diffusion models generate high-quality images through progressive denoising but are computationally intensive due to large model sizes and repeated sampling. Knowledge distillation, which transfers knowledge from a complex teacher to a simpler student model, has been widely studied in recognition tasks, particularly for transferring concepts unseen during student training. However, its application to diffusion models remains underexplored, especially in enabling student models to generate concepts not covered by the training images. In this work, we propose Random Conditioning, a novel approach that pairs noised images with randomly selected text conditions to enable efficient, image-free knowledge distillation. By leveraging this technique, we show that the student can generate concepts unseen in the training images. When applied to conditional diffusion model distillation, our method allows the student to explore the condition space without generating condition-specific images, resulting in notable improvements in both generation quality and efficiency. This promotes resource-efficient deployment of generative diffusion models, broadening their accessibility for both research and real-world applications. Code, models, and datasets are available at https://dohyun-as.github.io/Random-Conditioning .

cs.LG↗

arXiv2025

AudioGenX: Explainability on Text-to-Audio Generative Models

Text-to-audio generation models (TAG) have achieved significant advances in generating audio conditioned on text descriptions. However, a critical challenge lies in the lack of transparency regarding how each textual input impacts the generated audio. To address this issue, we introduce AudioGenX, an Explainable AI (XAI) method that provides explanations for text-to-audio generation models by highlighting the importance of input tokens. AudioGenX optimizes an Explainer by leveraging factual and counterfactual objective functions to provide faithful explanations at the audio token level. This method offers a detailed and comprehensive understanding of the relationship between text inputs and audio outputs, enhancing both the explainability and trustworthiness of TAG models. Extensive experiments demonstrate the effectiveness of AudioGenX in producing faithful explanations, benchmarked against existing methods using novel evaluation metrics specifically designed for audio generation tasks.

cs.SD↗

arXiv2025

Adaptive sequential Monte Carlo for structured cross validation in Bayesian hierarchical models

Importance sampling (IS) is commonly used for cross validation (CV) in Bayesian models, because it only involves reweighting existing posterior draws without needing to re-estimate the model by re-running Markov chain Monte Carlo (MCMC). For hierarchical models, standard IS can be unreliable; the out-of-sample generalization hypothesis may involve structured case-deletion schemes which significantly alter the posterior geometry. This can force costly MCMC re-runs and make CV impractical. As a principled alternative, we tailor adaptive sequential Monte Carlo to sample along a path of posteriors that leads to the case-deleted posterior. The sampler is designed to support various hypotheses by accommodating diverse CV designs, and to streamline the workflow by automating path construction and systematically minimizing MCMC intervention. We demonstrate its utility with three types of predictive model assessment: longitudinal leave-group-out CV, group $K$-fold CV, and sequential one-step-ahead validation.

stat.CO↗

arXiv2024

Zero-inflated stochastic volatility model for disaggregated inflation data with exact zeros

The disaggregated time-series for the Consumer Price Index (CPI) often exhibits exact zero price changes, stemming from structural features of the data collection process. However, the currently prominent stochastic volatility model of trend-inflation is designed for aggregate measures of price inflation, where zeros rarely occur. We formulate a zero-inflated stochastic volatility model applicable to such non-stationary, real-valued, multivariate time-series data with exact zeros, which jointly specifies the dynamic zero-generating process. For posterior inference, an efficient custom P\'olya--Gamma augmented Gibbs sampler is derived. Applying the model to disaggregated CPI data in four advanced economies -- US, UK, Germany, and Japan -- we find that the zero-inflated model yields more informative estimates of time-varying trend and volatility, as it accounts for the presence of zeros and avoids underestimation. In an out-of-sample forecasting exercise, we find that the zero-inflated model delivers improved point forecasts and better calibrated interval forecasts, particularly when zero-inflation is prevalent.

stat.ME↗