Donghyeok Shin

Donghyeok Shin, Seungjae Shin, Il-Chul Moon

This paper presents FreD, a novel parameterization method for dataset distillation, which utilizes the frequency domain to distill a small-sized synthetic dataset from a large-sized original dataset. Unlike conventional approaches that focus on the spatial domain, FreD employs frequency-based transforms to optimize the frequency representations of each data instance. By leveraging the concentration of spatial domain information on specific frequency components, FreD intelligently selects a subset of frequency dimensions for optimization, leading to a significant reduction in the required budget for synthesizing an instance. Through the selection of frequency dimensions based on the explained variance, FreD demonstrates both theoretical and empirical evidence of its ability to operate efficiently within a limited budget, while better preserving the information of the original dataset compared to conventional parameterization methods. Furthermore, based on the orthogonal compatibility of FreD with existing methods, we confirm that FreD consistently improves the performances of existing distillation methods over the evaluation scenarios with different benchmark datasets. We release the code at https://github.com/sdh0818/FreD.

JoonHo Jang, Byeonghu Na, DongHyeok Shin et al.

Open-Set Domain Adaptation (OSDA) assumes that a target domain contains unknown classes, which are not discovered in a source domain. Existing domain adversarial learning methods are not suitable for OSDA because distribution matching with $\textit{unknown}$ classes leads to negative transfer. Previous OSDA methods have focused on matching the source and the target distribution by only utilizing $\textit{known}$ classes. However, this $\textit{known}$-only matching may fail to learn the target-$\textit{unknown}$ feature space. Therefore, we propose Unknown-Aware Domain Adversarial Learning (UADAL), which $\textit{aligns}$ the source and the target-$\textit{known}$ distribution while simultaneously $\textit{segregating}$ the target-$\textit{unknown}$ distribution in the feature alignment procedure. We provide theoretical analyses on the optimized state of the proposed $\textit{unknown-aware}$ feature alignment, so we can guarantee both $\textit{alignment}$ and $\textit{segregation}$ theoretically. Empirically, we evaluate UADAL on the benchmark datasets, which shows that UADAL outperforms other methods with better feature alignments by reporting state-of-the-art performances.

Seungjae Shin, Heesun Bae, Donghyeok Shin et al.

Training neural networks on a large dataset requires substantial computational costs. Dataset reduction selects or synthesizes data instances based on the large dataset, while minimizing the degradation in generalization performance from the full dataset. Existing methods utilize the neural network during the dataset reduction procedure, so the model parameter becomes important factor in preserving the performance after reduction. By depending upon the importance of parameters, this paper introduces a new reduction objective, coined LCMat, which Matches the Loss Curvatures of the original dataset and reduced dataset over the model parameter space, more than the parameter point. This new objective induces a better adaptation of the reduced dataset on the perturbed parameter region than the exact point matching. Particularly, we identify the worst case of the loss curvature gap from the local parameter region, and we derive the implementable upper bound of such worst-case with theoretical analyses. Our experiments on both coreset selection and condensation benchmarks illustrate that LCMat shows better generalization performances than existing baselines.

Donghyeok Shin, HeeSun Bae, Gyuwon Sim et al.

Utilizing a large-scale dataset is essential for training high-performance deep learning models, but it also comes with substantial computation and storage costs. To overcome these challenges, dataset distillation has emerged as a promising solution by compressing the large-scale dataset into a smaller synthetic dataset that retains the essential information needed for training. This paper proposes a novel parameterization framework for dataset distillation, coined Distilling Dataset into Neural Field (DDiF), which leverages the neural field to store the necessary information of the large-scale dataset. Due to the unique nature of the neural field, which takes coordinates as input and output quantity, DDiF effectively preserves the information and easily generates various shapes of data. We theoretically confirm that DDiF exhibits greater expressiveness than some previous literature when the utilized budget for a single synthetic instance is the same. Through extensive experiments, we demonstrate that DDiF achieves superior performance on several benchmark datasets, extending beyond the image domain to include video, audio, and 3D voxel. We release the code at https://github.com/aailab-kaist/DDiF.

Yeongmin Kim, Donghyeok Shin, Byeonghu Na et al.

Diffusion models have demonstrated strong generative performance; however, generated samples often fail to fully align with human intent. This paper studies a test-time scaling method that enables sampling from regions with higher human-aligned reward values. Existing gradient guidance methods approximate the expected future reward (EFR) at an intermediate particle $\mathbf{x}_t$ using a Taylor approximation, but this approximation at each time step incurs high computational cost due to sequential neural backpropagation. We show that the EFR at any $\mathbf{x}_t$ can be computed using only marginal samples from a pre-trained diffusion model. The proposed EFR formulation detaches the neural dependency between $\mathbf{x}_t$ and the EFR, enabling closed-form guidance computation without neural backpropagation. To further improve efficiency, we introduce lookahead sampling to collect marginal samples. For final sample generation, we use an accurate solver that guides particles toward high-reward lookahead samples. We refer to this sampling scheme as LiDAR sampling. LiDAR achieves substantial performance improvements using only three samples with a 3-step lookahead solver, exhibiting steep performance gains as lookahead accuracy and sample count increase; notably, it reaches the same GenEval performance as the latest gradient guidance method for SDXL with a 9.5x speedup.

Byeonghu Na, Minsang Park, Gyuwon Sim et al.

Text-to-image diffusion models rely on text embeddings from a pre-trained text encoder, but these embeddings remain fixed across all diffusion timesteps, limiting their adaptability to the generative process. We propose Diffusion Adaptive Text Embedding (DATE), which dynamically updates text embeddings at each diffusion timestep based on intermediate perturbed data. We formulate an optimization problem and derive an update rule that refines the text embeddings at each sampling step to improve alignment and preference between the mean predicted image and the text. This allows DATE to dynamically adapts the text conditions to the reverse-diffused images throughout diffusion sampling without requiring additional model training. Through theoretical analysis and empirical results, we show that DATE maintains the generative capability of the model while providing superior text-image alignment over fixed text embeddings across various tasks, including multi-concept generation and text-guided image editing. Our code is available at https://github.com/aailab-kaist/DATE.

Yeongmin Kim, Donghyeok Shin, Mina Kang et al.

Large language models (LLMs) deliver remarkable performance but are costly to deploy, motivating knowledge distillation (KD) for efficient inference. Existing KD objectives typically match student and teacher probabilities via softmax, which blurs valuable logit information. While direct logit distillation (DLD) mitigates softmax smoothing, it fails to account for logit shift invariance, thereby restricting the solution space. We propose Concrete Score Distillation (CSD), a discrete score-matching objective that overcomes both softmax-induced smoothing and restrictions on the optimal solution set. We resolve the training instability and quadratic complexity of discrete score-matching in autoregressive LLMs, and the resulting CSD objective aligns relative logit differences across all vocabulary pairs between student and teacher with flexible weighting. We provide both mode-seeking and mode-covering instances within our framework and evaluate CSD on task-agnostic instruction-following and task-specific distillation using GPT-2-1.5B, OpenLLaMA-7B, and GEMMA-7B-IT. Experiments show that CSD consistently surpasses recent KD objectives, achieves favorable fidelity-diversity trade-offs, and yields complementary gains when combined with on-policy techniques, demonstrating its scalability and effectiveness for LLM distillation.

Donghyeok Shin, Yeongmin Kim, Suhyeon Jo et al.

Autoregressive large language models (LLMs) have achieved remarkable improvement across many tasks but incur high computational and memory costs. Knowledge distillation (KD) mitigates this issue by transferring knowledge from a large teacher to a smaller student through distributional alignment. Previous studies have proposed various discrepancy metrics, but the capacity gap and training instability caused by near-zero probabilities, stemming from the high-dimensional output of LLMs, remain fundamental limitations. To overcome these challenges, several approaches implicitly or explicitly incorporating assistant distribution have recently been proposed. However, the past proposals of assistant distributions have been a fragmented approach without a systematic investigation of the interpolation path and the divergence. This paper proposes $α$-mixture assistant distribution, a novel generalized family of assistant distributions, and $α$-mixture distillation, coined AMiD, a unified framework for KD using the assistant distribution. The $α$-mixture assistant distribution provides a continuous extension of the assistant distribution by introducing a new distribution design variable $α$, which has been fixed in all previous approaches. Furthermore, AMiD generalizes the family of divergences used with the assistant distributions based on optimality, which has also been restricted in previous works. Through extensive experiments, we demonstrate that AMiD offers superior performance and training stability by leveraging a broader and theoretically grounded assistant distribution space.