Eunhyeok Park
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Eunhyeok Park has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Changhun Lee
Department of Convergence IT Engineering
7 shared publications Jun-Gyu Jin
Graduate School of Artificial Intelligence
5 shared publicationsJungyu Jin
Graduate School of Artificial Intelligence Pohang University of Science and Technology (POSTECH
4 shared publications- Junhyuk So
Department of Computer Science and Engineering POSTECH
4 shared publications Byeongho Yu
3 shared publications- Gwangsun Kim
POSTECH, Pohang, Republic of Korea
3 shared publications
Latest Publications
- Open Access
Stabilizing Direct Training of Spiking Neural Networks: Membrane Potential Initialization and Threshold-robust Surrogate Gradient
Recent advancements in the direct training of Spiking Neural Networks (SNNs) have demonstrated high-quality outputs even at early timesteps, paving the way for novel energy-efficient AI paradigms. However, the inherent non-linearity and temporal dependencies in SNNs introduce persistent challenges, such as temporal covariate shift (TCS) and unstable gradient flow with learnable neuron thresholds. In this paper, we present two key innovations: MP-Init (Membrane Potential Initialization) and TrSG (Threshold-robust Surrogate Gradient). MP-Init addresses TCS by aligning the initial membrane potential with its stationary distribution, while TrSG stabilizes gradient flow with respect to threshold voltage during training. Extensive experiments validate our approach, achieving state-of-the-art accuracy on both static and dynamic image datasets. The code is available at: https://github.com/kookhh0827/SNN-MP-Init-TRSG
Open AccessStabilizing Direct Training of Spiking Neural Networks: Membrane Potential Initialization and Threshold-robust Surrogate Gradient
Recent advancements in the direct training of Spiking Neural Networks (SNNs) have demonstrated high-quality outputs even at early timesteps, paving the way for novel energy-efficient AI paradigms. However, the inherent non-linearity and temporal dependencies in SNNs introduce persistent challenges, such as temporal covariate shift (TCS) and unstable gradient flow with learnable neuron thresholds. In this paper, we present two key innovations: MP-Init (Membrane Potential Initialization) and TrSG (Threshold-robust Surrogate Gradient). MP-Init addresses TCS by aligning the initial membrane potential with its stationary distribution, while TrSG stabilizes gradient flow with respect to threshold voltage during training. Extensive experiments validate our approach, achieving state-of-the-art accuracy on both static and dynamic image datasets. The code is available at: https://github.com/kookhh0827/SNN-MP-Init-TRSG
- Open Access
Improving Generative Behavior Cloning via Self-Guidance and Adaptive Chunking
Generative Behavior Cloning (GBC) is a simple yet effective framework for robot learning, particularly in multi-task settings. Recent GBC methods often employ diffusion policies with open-loop (OL) control, where actions are generated via a diffusion process and executed in multi-step chunks without replanning. While this approach has demonstrated strong success rates and generalization, its inherent stochasticity can result in erroneous action sampling, occasionally leading to unexpected task failures. Moreover, OL control suffers from delayed responses, which can degrade performance in noisy or dynamic environments. To address these limitations, we propose two novel techniques to enhance the consistency and reactivity of diffusion policies: (1) self-guidance, which improves action fidelity by leveraging past observations and implicitly promoting future-aware behavior; and (2) adaptive chunking, which selectively updates action sequences when the benefits of reactivity outweigh the need for temporal consistency. Extensive experiments show that our approach substantially improves GBC performance across a wide range of simulated and real-world robotic manipulation tasks. Our code is available at https://github.com/junhyukso/SGAC
Open AccessImproving Generative Behavior Cloning via Self-Guidance and Adaptive Chunking
Generative Behavior Cloning (GBC) is a simple yet effective framework for robot learning, particularly in multi-task settings. Recent GBC methods often employ diffusion policies with open-loop (OL) control, where actions are generated via a diffusion process and executed in multi-step chunks without replanning. While this approach has demonstrated strong success rates and generalization, its inherent stochasticity can result in erroneous action sampling, occasionally leading to unexpected task failures. Moreover, OL control suffers from delayed responses, which can degrade performance in noisy or dynamic environments. To address these limitations, we propose two novel techniques to enhance the consistency and reactivity of diffusion policies: (1) self-guidance, which improves action fidelity by leveraging past observations and implicitly promoting future-aware behavior; and (2) adaptive chunking, which selectively updates action sequences when the benefits of reactivity outweigh the need for temporal consistency. Extensive experiments show that our approach substantially improves GBC performance across a wide range of simulated and real-world robotic manipulation tasks. Our code is available at https://github.com/junhyukso/SGAC
- Open Access
PruneCD: Contrasting Pruned Self Model to Improve Decoding Factuality
To mitigate the hallucination problem in large language models, DoLa exploits early exit logits from the same model as a contrastive prior. However, we found that these early exit logits tend to be flat, low in magnitude, and fail to reflect meaningful contrasts. To address this, we propose PruneCD, a novel contrastive decoding method that constructs the amateur model via layer pruning rather than early exit. This design leads to more informative and well-aligned logits, enabling more effective contrastive decoding. Through qualitative and quantitative analyses, we demonstrate that PruneCD consistently improves factuality with minimal inference overhead, offering a robust and practical approach to mitigating hallucinations in LLMs.
Open AccessPruneCD: Contrasting Pruned Self Model to Improve Decoding Factuality
To mitigate the hallucination problem in large language models, DoLa exploits early exit logits from the same model as a contrastive prior. However, we found that these early exit logits tend to be flat, low in magnitude, and fail to reflect meaningful contrasts. To address this, we propose PruneCD, a novel contrastive decoding method that constructs the amateur model via layer pruning rather than early exit. This design leads to more informative and well-aligned logits, enabling more effective contrastive decoding. Through qualitative and quantitative analyses, we demonstrate that PruneCD consistently improves factuality with minimal inference overhead, offering a robust and practical approach to mitigating hallucinations in LLMs.
- Open Access
AMQ: Enabling AutoML for Mixed-precision Weight-Only Quantization of Large Language Models
To enable broader deployment of Large Language Models (LLMs), it is essential to identify the best-performing model under strict memory constraints. We present AMQ, Automated Mixed-Precision Weight-Only Quantization, a framework that assigns layer-wise quantization bit-widths to optimally balance model quality and memory usage. However, the combinatorial search space, with over 10^{100} possible configurations, makes conventional black-box optimization infeasible. AMQ overcomes this challenge through four key innovations:(1) search space pruning using prior knowledge to exclude unpromising configurations, (2) quantization proxy to bypass costly format conversions during search, (3) quality predictor to minimize evaluation overhead, and (4) iterative search-and-update strategy for fast and stable convergence. By integrating these components, AMQ efficiently explores the quality-efficiency landscape, reaching the Pareto frontier and yielding LLMs that are both compact and high-performing. Our code is available at https://github.com/dlwns147/amq.
Open AccessAMQ: Enabling AutoML for Mixed-precision Weight-Only Quantization of Large Language Models
To enable broader deployment of Large Language Models (LLMs), it is essential to identify the best-performing model under strict memory constraints. We present AMQ, Automated Mixed-Precision Weight-Only Quantization, a framework that assigns layer-wise quantization bit-widths to optimally balance model quality and memory usage. However, the combinatorial search space, with over 10^{100} possible configurations, makes conventional black-box optimization infeasible. AMQ overcomes this challenge through four key innovations:(1) search space pruning using prior knowledge to exclude unpromising configurations, (2) quantization proxy to bypass costly format conversions during search, (3) quality predictor to minimize evaluation overhead, and (4) iterative search-and-update strategy for fast and stable convergence. By integrating these components, AMQ efficiently explores the quality-efficiency landscape, reaching the Pareto frontier and yielding LLMs that are both compact and high-performing. Our code is available at https://github.com/dlwns147/amq.
- Open Access
Grouped Speculative Decoding for Autoregressive Image Generation
Recently, autoregressive (AR) image models have demonstrated remarkable generative capabilities, positioning themselves as a compelling alternative to diffusion models. However, their sequential nature leads to long inference times, limiting their practical scalability. In this work, we introduce Grouped Speculative Decoding (GSD), a novel, training-free acceleration method for AR image models. While recent studies have explored Speculative Decoding (SD) as a means to speed up AR image generation, existing approaches either provide only modest acceleration or require additional training. Our in-depth analysis reveals a fundamental difference between language and image tokens: image tokens exhibit inherent redundancy and diversity, meaning multiple tokens can convey valid semantics. However, traditional SD methods are designed to accept only a single most-likely token, which fails to leverage this difference, leading to excessive false-negative rejections. To address this, we propose a new SD strategy that evaluates clusters of visually valid tokens rather than relying on a single target token. Additionally, we observe that static clustering based on embedding distance is ineffective, which motivates our dynamic GSD approach. Extensive experiments show that GSD accelerates AR image models by an average of 3.7x while preserving image quality-all without requiring any additional training. The source code is available at https://github.com/junhyukso/GSD
Open AccessGrouped Speculative Decoding for Autoregressive Image Generation
Recently, autoregressive (AR) image models have demonstrated remarkable generative capabilities, positioning themselves as a compelling alternative to diffusion models. However, their sequential nature leads to long inference times, limiting their practical scalability. In this work, we introduce Grouped Speculative Decoding (GSD), a novel, training-free acceleration method for AR image models. While recent studies have explored Speculative Decoding (SD) as a means to speed up AR image generation, existing approaches either provide only modest acceleration or require additional training. Our in-depth analysis reveals a fundamental difference between language and image tokens: image tokens exhibit inherent redundancy and diversity, meaning multiple tokens can convey valid semantics. However, traditional SD methods are designed to accept only a single most-likely token, which fails to leverage this difference, leading to excessive false-negative rejections. To address this, we propose a new SD strategy that evaluates clusters of visually valid tokens rather than relying on a single target token. Additionally, we observe that static clustering based on embedding distance is ineffective, which motivates our dynamic GSD approach. Extensive experiments show that GSD accelerates AR image models by an average of 3.7x while preserving image quality-all without requiring any additional training. The source code is available at https://github.com/junhyukso/GSD
- Open Access
The Oversmoothing Fallacy: A Misguided Narrative in GNN Research
Oversmoothing has been recognized as a main obstacle to building deep Graph Neural Networks (GNNs), limiting the performance. This position paper argues that the influence of oversmoothing has been overstated and advocates for a further exploration of deep GNN architectures. Given the three core operations of GNNs, aggregation, linear transformation, and non-linear activation, we show that prior studies have mistakenly confused oversmoothing with the vanishing gradient, caused by transformation and activation rather than aggregation. Our finding challenges prior beliefs about oversmoothing being unique to GNNs. Furthermore, we demonstrate that classical solutions such as skip connections and normalization enable the successful stacking of deep GNN layers without performance degradation. Our results clarify misconceptions about oversmoothing and shed new light on the potential of deep GNNs.
Open AccessThe Oversmoothing Fallacy: A Misguided Narrative in GNN Research
Oversmoothing has been recognized as a main obstacle to building deep Graph Neural Networks (GNNs), limiting the performance. This position paper argues that the influence of oversmoothing has been overstated and advocates for a further exploration of deep GNN architectures. Given the three core operations of GNNs, aggregation, linear transformation, and non-linear activation, we show that prior studies have mistakenly confused oversmoothing with the vanishing gradient, caused by transformation and activation rather than aggregation. Our finding challenges prior beliefs about oversmoothing being unique to GNNs. Furthermore, we demonstrate that classical solutions such as skip connections and normalization enable the successful stacking of deep GNN layers without performance degradation. Our results clarify misconceptions about oversmoothing and shed new light on the potential of deep GNNs.
- Open Access
Efficient Latent Semantic Clustering for Scaling Test-Time Computation of LLMs
Scaling test-time computation--generating and analyzing multiple or sequential outputs for a single input--has become a promising strategy for improving the reliability and quality of large language models (LLMs), as evidenced by advances in uncertainty quantification and multi-step reasoning. A key shared component is semantic clustering, which groups outputs that differ in form but convey the same meaning. Semantic clustering enables estimation of the distribution over the semantics of outputs and helps avoid redundant exploration of reasoning paths. However, existing approaches typically rely on external models, which introduce substantial computational overhead and often fail to capture context-aware semantics. We propose Latent Semantic Clustering (LSC), a lightweight and context-sensitive method that leverages the generator LLM's internal hidden states for clustering, eliminating the need for external models. Our extensive experiment across various LLMs and datasets shows that LSC significantly improves the computational efficiency of test-time scaling while maintaining or exceeding the performance of existing methods.
Open AccessEfficient Latent Semantic Clustering for Scaling Test-Time Computation of LLMs
Scaling test-time computation--generating and analyzing multiple or sequential outputs for a single input--has become a promising strategy for improving the reliability and quality of large language models (LLMs), as evidenced by advances in uncertainty quantification and multi-step reasoning. A key shared component is semantic clustering, which groups outputs that differ in form but convey the same meaning. Semantic clustering enables estimation of the distribution over the semantics of outputs and helps avoid redundant exploration of reasoning paths. However, existing approaches typically rely on external models, which introduce substantial computational overhead and often fail to capture context-aware semantics. We propose Latent Semantic Clustering (LSC), a lightweight and context-sensitive method that leverages the generator LLM's internal hidden states for clustering, eliminating the need for external models. Our extensive experiment across various LLMs and datasets shows that LSC significantly improves the computational efficiency of test-time scaling while maintaining or exceeding the performance of existing methods.
- Open Access
HOT: Hadamard-based Optimized Training
It has become increasingly important to optimize backpropagation to reduce memory usage and computational overhead. Achieving this goal is highly challenging, as multiple objectives must be considered jointly while maintaining training quality. In this paper, we focus on matrix multiplication, which accounts for the largest portion of training costs, and analyze its backpropagation in detail to identify lightweight techniques that offer the best benefits. Based on this analysis, we introduce a novel method, Hadamard-based Optimized Training (HOT). In this approach, we apply Hadamard-based optimizations, such as Hadamard quantization and Hadamard low-rank approximation, selectively and with awareness of the suitability of each optimization for different backward paths. Additionally, we introduce two enhancements: activation buffer compression and layer-wise quantizer selection. Our extensive analysis shows that HOT achieves up to 75% memory savings and a 2.6 times acceleration on real GPUs, with negligible accuracy loss compared to FP32 precision.
Open AccessHOT: Hadamard-based Optimized Training
It has become increasingly important to optimize backpropagation to reduce memory usage and computational overhead. Achieving this goal is highly challenging, as multiple objectives must be considered jointly while maintaining training quality. In this paper, we focus on matrix multiplication, which accounts for the largest portion of training costs, and analyze its backpropagation in detail to identify lightweight techniques that offer the best benefits. Based on this analysis, we introduce a novel method, Hadamard-based Optimized Training (HOT). In this approach, we apply Hadamard-based optimizations, such as Hadamard quantization and Hadamard low-rank approximation, selectively and with awareness of the suitability of each optimization for different backward paths. Additionally, we introduce two enhancements: activation buffer compression and layer-wise quantizer selection. Our extensive analysis shows that HOT achieves up to 75% memory savings and a 2.6 times acceleration on real GPUs, with negligible accuracy loss compared to FP32 precision.
- Open Access
PCM : Picard Consistency Model for Fast Parallel Sampling of Diffusion Models
Recently, diffusion models have achieved significant advances in vision, text, and robotics. However, they still face slow generation speeds due to sequential denoising processes. To address this, a parallel sampling method based on Picard iteration was introduced, effectively reducing sequential steps while ensuring exact convergence to the original output. Nonetheless, Picard iteration does not guarantee faster convergence, which can still result in slow generation in practice. In this work, we propose a new parallelization scheme, the Picard Consistency Model (PCM), which significantly reduces the number of generation steps in Picard iteration. Inspired by the consistency model, PCM is directly trained to predict the fixed-point solution, or the final output, at any stage of the convergence trajectory. Additionally, we introduce a new concept called model switching, which addresses PCM's limitations and ensures exact convergence. Extensive experiments demonstrate that PCM achieves up to a 2.71x speedup over sequential sampling and a 1.77x speedup over Picard iteration across various tasks, including image generation and robotic control.
Open AccessPCM : Picard Consistency Model for Fast Parallel Sampling of Diffusion Models
Recently, diffusion models have achieved significant advances in vision, text, and robotics. However, they still face slow generation speeds due to sequential denoising processes. To address this, a parallel sampling method based on Picard iteration was introduced, effectively reducing sequential steps while ensuring exact convergence to the original output. Nonetheless, Picard iteration does not guarantee faster convergence, which can still result in slow generation in practice. In this work, we propose a new parallelization scheme, the Picard Consistency Model (PCM), which significantly reduces the number of generation steps in Picard iteration. Inspired by the consistency model, PCM is directly trained to predict the fixed-point solution, or the final output, at any stage of the convergence trajectory. Additionally, we introduce a new concept called model switching, which addresses PCM's limitations and ensures exact convergence. Extensive experiments demonstrate that PCM achieves up to a 2.71x speedup over sequential sampling and a 1.77x speedup over Picard iteration across various tasks, including image generation and robotic control.
- Open Access
- 13 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 24 (1) , 53-56
Cost-Effective Extension of DRAM-PIM for Group-Wise LLM Quantization
Processing-in-Memory (PIM) is emerging as a promising next-generation hardware to address memory bottlenecks in large language model (LLM) inference by leveraging internal memory bandwidth, enabling more energy-efficient on-device AI. However, LLMs’ large footprint poses significant challenges for accelerating them on PIM due to limited available space. Recent advances in weight-only quantization, especially group-wise weight quantization (GWQ), reduce LLM model sizes, enabling parameters to be stored at 4-bit precision or lower with minimal accuracy loss. Despite this, current PIM architectures experience performance degradation when handling the additional computations required for quantized weights. While incorporating extra logic could mitigate this degradation, it is often prohibitively expensive due to the constraints of memory technology, necessitating solutions with minimal area overhead. This work introduces two key innovations: 1) scale cascading, and 2) an INT2FP converter, to support GWQ-applied LLMs on PIM with minimal dequantization latency and area overhead compared to FP16 GEMV. Experimental results show that the proposed approach adds less than 0.6% area overhead to the existing PIM unit and achieves a 7% latency overhead for dequantization and GEMV in 4-bit GWQ with a group size of 128, compared to FP16 GEMV, while offering a 1.55× performance gain over baseline dequantization.
- 13 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 24 (1) , 53-56
Open AccessCost-Effective Extension of DRAM-PIM for Group-Wise LLM Quantization
Processing-in-Memory (PIM) is emerging as a promising next-generation hardware to address memory bottlenecks in large language model (LLM) inference by leveraging internal memory bandwidth, enabling more energy-efficient on-device AI. However, LLMs’ large footprint poses significant challenges for accelerating them on PIM due to limited available space. Recent advances in weight-only quantization, especially group-wise weight quantization (GWQ), reduce LLM model sizes, enabling parameters to be stored at 4-bit precision or lower with minimal accuracy loss. Despite this, current PIM architectures experience performance degradation when handling the additional computations required for quantized weights. While incorporating extra logic could mitigate this degradation, it is often prohibitively expensive due to the constraints of memory technology, necessitating solutions with minimal area overhead. This work introduces two key innovations: 1) scale cascading, and 2) an INT2FP converter, to support GWQ-applied LLMs on PIM with minimal dequantization latency and area overhead compared to FP16 GEMV. Experimental results show that the proposed approach adds less than 0.6% area overhead to the existing PIM unit and achieves a 7% latency overhead for dequantization and GEMV in 4-bit GWQ with a group size of 128, compared to FP16 GEMV, while offering a 1.55× performance gain over baseline dequantization.