arXiv subjects

Jaemin Kim

Publications and source records attributed to Jaemin Kim.

At least 37 records · Page 2Linked to original sources

arXiv2026

Speculative Verification: Exploiting Information Gain to Refine Speculative Decoding

LLMs have low GPU efficiency and high latency due to autoregressive decoding. Speculative decoding (SD) mitigates this using a small draft model to speculatively generate multiple tokens, which are then verified in parallel by a target model. However, when speculation accuracy is low, the overhead from rejected tokens can offset the benefits, limiting SD's effectiveness, especially at large batch sizes. To address this, we propose Speculative Verification (SV), an efficient augmentation to SD that dynamically predicts speculation accuracy and adapts the verification length to maximize throughput. SV introduces a companion model - a small auxiliary model similar in size to the draft model - to estimate the alignment between draft and target model distributions. By maximizing the information gain from quantifying this alignment, SV refines verification decisions, reducing wasted computation on rejected tokens and improving decoding efficiency. Moreover, SV requires no modifications to the draft or target models and is compatible with existing SD variants. We extensively evaluated SV on publicly available LLMs across three NLP tasks using nine combinations of draft, companion, and target models, including 13B-72B target models and three types of variations: base (no finetuning), instruction-tuned, and task fine-tuned. Across all experiments and batch sizes (4-80), SV consistently outperforms both SD and standard decoding with the target model. It improves SD performance by up to 2$\times$, with an average speedup of 1.4 $\times$ in large-batch settings (batch sizes 32-80). These results demonstrate SV's robustness, scalability, and practical utility for efficient LLM inference.

cs.CL↗

arXiv2026

On the complementary roles of anisotropic crack density and anisotropic crack driving force in phase-field modeling of mixed-mode fracture

Phase-field models for anisotropic fracture employ two complementary mechanisms: (i) the anisotropic crack density function, controlling direction-dependent fracture resistance, and (ii) the anisotropic strain energy, governing the fracture driving force. Although the unified framework was presented in Pranavi et al.[Comput. Mech., 73 (2024)], the distinct roles of these mechanisms and their interaction remain uninvestigated. This work addresses this gap by first validating the formulation against mixed-mode fracture experiments on a soft elastomer (Lu et al. [Extreme Mech. Lett., 48 (2021)]), and then conducting systematic parametric studies on single-edge-notched (SEN) and open-hole tension (OHT) specimens to isolate each mechanism. The SEN studies show that the crack density anisotropy controls the crack path and toughness while leaving the elastic response unchanged, whereas the anisotropic strain energy deflects the crack but saturates rapidly. The OHT studies reveal a geometry-dependent role expansion: the anisotropic strain energy governs fiber-orientation-dependent stiffness, peak force, and fracture displacement. When both mechanisms act together, the combined response exhibits nonlinear synergistic interaction exceeding the linear sum of the individual contributions. These results establish that the crack density anisotropy governs the crack path (fracture resistance), while the anisotropic strain energy governs the driving force and, in stress-concentration geometries, additionally controls the elastic strain energy distribution around the stress concentrator.

cond-mat.mtrl-sci↗

arXiv2026

Adaptive Guidance for Retrieval-Augmented Masked Diffusion Models

Retrieval-Augmented Generation (RAG) improves factual grounding by incorporating external knowledge into language model generation. However, when retrieved context is noisy, unreliable, or inconsistent with the model's parametric knowledge, it introduces retrieval-prior conflicts that can degrade generation quality. While this problem has been studied in autoregressive language models, it remains largely unexplored in diffusion-based language models, where the iterative denoising process introduces unique challenges for integrating retrieved context. In this work, we propose Adaptive Retrieval-Augmented Masked Diffusion (ARAM), a training-free adaptive guidance framework for Masked Diffusion Models (MDMs) in RAG settings. ARAM dynamically calibrates the guidance scale during denoising according to the Signal-to-Noise Ratio (SNR) of the distributional shift induced by retrieved context. Intuitively, the model strengthens guidance when the retrieved context provides reliable corrective evidence and suppresses it when the contextual signal is noisy or non-supportive. Extensive experiments on multiple knowledge-intensive QA benchmarks show that ARAM improves overall QA performance over competitive RAG baselines.

cs.CL↗

arXiv2026

CLaD: Planning with Grounded Foresight via Cross-Modal Latent Dynamics

Robotic manipulation involves kinematic and semantic transitions that are inherently coupled via underlying actions. However, existing approaches plan within either semantic or latent space without explicitly aligning these cross-modal transitions. To address this, we propose CLaD, a framework that models how proprioceptive and semantic states jointly evolve under actions through asymmetric cross-attention that allows kinematic transitions to query semantic ones. CLaD predicts grounded latent foresights via self-supervised objectives with EMA target encoders and auxiliary reconstruction losses, preventing representation collapse while anchoring predictions to observable states. Predicted foresights are modulated with observations to condition a diffusion policy for action generation. On LIBERO-LONG benchmark, CLaD achieves 94.7\% success rate, competitive with large VLAs with significantly fewer parameters.

cs.RO↗

arXiv2026

Dementia-R1: Reinforced Pretraining and Reasoning from Unstructured Clinical Notes for Real-World Dementia Prognosis

While Large Language Models (LLMs) have shown strong performance on clinical text understanding, they struggle with longitudinal prediction tasks such as dementia prognosis, which require reasoning over complex, non-monotonic symptom trajectories across multiple visits. Standard supervised training lacks explicit annotations for symptom evolution, while direct Reinforcement Learning (RL) is hindered by sparse binary rewards. To address this challenge, we introduce Dementia-R1, an RL-based framework for longitudinal dementia prognosis from unstructured clinical notes. Our approach adopts a Cold-Start RL strategy that pre-trains the model to predict verifiable clinical indices extracted from patient histories, enhancing the capability to reason about disease progression before determining the final clinical status. Extensive experiments show that Dementia-R1 achieves the best overall performance on the AMC real-world unstructured cohort, reaching an AUROC of 84.02% and outperforming models up to 10x larger. The framework also generalizes to Parkinson's disease dementia prediction in an independent hospital cohort, achieving an AUROC of 78.37%. On the ADNI benchmark, our 7B model attains the highest AUROC among all LLM baselines at 83.17%, demonstrating strong longitudinal reasoning over fluctuating cognitive trajectories. Code is available at https://anonymous.4open.science/r/dementiar1-CDB5.

cs.CL↗

arXiv2026

DMD-augmented Unpaired Neural Schrödinger Bridge for Ultra-Low Field MRI Enhancement

Ultra Low Field (64 mT) brain MRI improves accessibility but suffers from reduced image quality compared to 3 T. As paired 64 mT - 3 T scans are scarce, we propose an unpaired 64 mT $\rightarrow$ 3 T translation framework that enhances realism while preserving anatomy. Our method builds upon the Unpaired Neural Schrödinge Bridge (UNSB) with multi-step refinement. To strengthen target distribution alignment, we augment the adversarial objective with DMD2-style diffusion-guided distribution matching using a frozen 3T diffusion teacher. To explicitly constrain global structure beyond patch-level correspondence, we combine PatchNCE with an Anatomical Structure Preservation (ASP) regularizer that enforces soft foreground background consistency and boundary aware constraints. Evaluated on two disjoint cohorts, the proposed framework achieves an improved realism structure trade-off, enhancing distribution level realism on unpaired benchmarks while increasing structural fidelity on the paired cohort compared to unpaired baselines.

cs.CV↗

arXiv2026

A unified variational framework for phase-field fracture and third-medium contact in finite deformation hyperelasticity

This paper presents a unified variational framework that integrates phase-field fracture (PFF) and third-medium contact (TMC) within finite deformation hyperelasticity. The key idea is that both crack and contact are treated through regularization: the sharp crack topology is regularized into a diffuse damage field, while the discrete contact interface is regularized by a compliant fictitious medium with auxiliary fields. This strategy eliminates the need for explicit contact detection or crack tracking algorithms. The framework is validated through two-dimensional three-point bending and three-dimensional Brazilian disk test simulations, demonstrating the interplay between contact-induced stress concentration and crack nucleation/propagation. In particular, the Brazilian disk simulation naturally reproduces secondary crushing-type fracture zones near the contact regions -- a phenomenon consistently observed in experiments yet inaccessible to simplified loading models. These results pave the way for predictive simulation of coupled contact-fracture phenomena without recourse to explicit interface tracking.

physics.comp-ph↗

arXiv2025

Free$^2$Guide: Training-Free Text-to-Video Alignment using Image LVLM

Diffusion models have achieved impressive results in generative tasks for text-to-video (T2V) synthesis. However, achieving accurate text alignment in T2V generation remains challenging due to the complex temporal dependencies across frames. Existing reinforcement learning (RL)-based approaches to enhance text alignment often require differentiable reward functions trained for videos, hindering their scalability and applicability. In this paper, we propose \textbf{Free$^2$Guide}, a novel gradient-free and training-free framework for aligning generated videos with text prompts. Specifically, leveraging principles from path integral control, Free$^2$Guide approximates guidance for diffusion models using non-differentiable reward functions, thereby enabling the integration of powerful black-box Large Vision-Language Models (LVLMs) as reward models. To enable image-trained LVLMs to assess text-to-video alignment, we leverage \textit{stitching} between video frames and use system prompts to capture sequential attributions. Our framework supports the flexible ensembling of multiple reward models to synergistically enhance alignment without significant computational overhead. Experimental results confirm that Free$^2$Guide using image-trained LVLMs significantly improves text-to-video alignment, thereby enhancing the overall video quality. Our results and code are available at https://kjm981995.github.io/free2guide/

cs.CV↗

arXiv2025

FlexiQ: Adaptive Mixed-Precision Quantization for Latency/Accuracy Trade-Offs in Deep Neural Networks

Neural networks commonly execute on hardware accelerators such as NPUs and GPUs for their size and computation overhead. These accelerators are costly and it is hard to scale their resources to handle real-time workload fluctuations. We present FlexiQ, an adaptive mixed-precision quantization scheme for computer vision models. FlexiQ selectively applies low-bitwidth computation to feature channels with small value ranges and employs an efficient bit-lowering method to minimize quantization errors while maintaining inference accuracy. Furthermore, FlexiQ adjusts its low-bitwidth channel ratio in real time, enabling quantized models to effectively manage fluctuating inference workload. We implemented FlexiQ prototype, including the mixed-precision inference runtime on our custom NPU and GPUs. Evaluated on eleven convolution- and transformer-based vision models, FlexiQ achieves on average 6.6% higher accuracy for 4-bit models with finetuning and outperforms four state-of-the-art quantization techniques. Moreover, our mixed-precision models achieved an efficient accuracy-latency trade-off, with the 50% 4-bit model incurring only 0.6% accuracy loss while achieving 40% of the speedup of the 100% 4-bit model over 8-bit model. Latency evaluations on our NPU and GPUs confirmed that FlexiQ introduces minimal runtime overhead, demonstrating its hardware efficiency and overall performance benefits.

cs.LG↗

arXiv2025

Generalized Consistency Trajectory Models for Image Manipulation

Diffusion models (DMs) excel in unconditional generation, as well as on applications such as image editing and restoration. The success of DMs lies in the iterative nature of diffusion: diffusion breaks down the complex process of mapping noise to data into a sequence of simple denoising tasks. Moreover, we are able to exert fine-grained control over the generation process by injecting guidance terms into each denoising step. However, the iterative process is also computationally intensive, often taking from tens up to thousands of function evaluations. Although consistency trajectory models (CTMs) enable traversal between any time points along the probability flow ODE (PFODE) and score inference with a single function evaluation, CTMs only allow translation from Gaussian noise to data. This work aims to unlock the full potential of CTMs by proposing generalized CTMs (GCTMs), which translate between arbitrary distributions via ODEs. We discuss the design space of GCTMs and demonstrate their efficacy in various image manipulation tasks such as image-to-image translation, restoration, and editing.

cs.CV↗

arXiv2025

Derivative-Free Diffusion Manifold-Constrained Gradient for Unified XAI

Gradient-based methods are a prototypical family of explainability techniques, especially for image-based models. Nonetheless, they have several shortcomings in that they (1) require white-box access to models, (2) are vulnerable to adversarial attacks, and (3) produce attributions that lie off the image manifold, leading to explanations that are not actually faithful to the model and do not align well with human perception. To overcome these challenges, we introduce Derivative-Free Diffusion Manifold-Constrainted Gradients (FreeMCG), a novel method that serves as an improved basis for explainability of a given neural network than the traditional gradient. Specifically, by leveraging ensemble Kalman filters and diffusion models, we derive a derivative-free approximation of the model's gradient projected onto the data manifold, requiring access only to the model's outputs. We demonstrate the effectiveness of FreeMCG by applying it to both counterfactual generation and feature attribution, which have traditionally been treated as distinct tasks. Through comprehensive evaluation on both tasks, counterfactual explanation and feature attribution, we show that our method yields state-of-the-art results while preserving the essential properties expected of XAI tools.

cs.CV↗

arXiv2025

Multiphysics Modeling of Surface Diffusion Coupled with Large Deformation in 3D Solids

We present a comprehensive theoretical and computational model that explores the behavior of a thin hydrated film bonded to a non-hydrated / impermeable soft substrate in the context of surface and bulk elasticity coupled with surface diffusion kinetics. This type of coupling can manifests as an integral aspect in diverse engineering processes encountered in optical interference coatings, tissue engineering, soft electronics, and can prove important in design process for the next generation of sensors and actuators, especially as the focus is shifted to systems in smaller lengthscales. The intricate interplay between solvent diffusion and deformation of the film is governed by surface poroelasticity, and the viscoelastic deformation of the substrate. While existing methodologies offer tools for studying coupled poroelasticity involving solvent diffusion and network deformation, there exists a gap in understanding how coupled poroelastic processes occurring in a film attached to the boundary of a highly deformable solid can influence its response. In this study, we introduce a non-equilibrium thermodynamics formulation encompassing the multiphysical processes of surface poroelasticity and bulk viscoelasticity, complemented by a corresponding finite element implementation. Our approach captures the complex dynamics between the finite deformation of the substrate and solvent diffusion on the surface. This work contributes valuable insights, particularly in scenarios where the coupling of surface diffusion kinetics and substrate elasticity is an important design factor.

cond-mat.soft↗

arXiv2025

Optical-Flow Guided Prompt Optimization for Coherent Video Generation

While text-to-video diffusion models have made significant strides, many still face challenges in generating videos with temporal consistency. Within diffusion frameworks, guidance techniques have proven effective in enhancing output quality during inference; however, applying these methods to video diffusion models introduces additional complexity of handling computations across entire sequences. To address this, we propose a novel framework called MotionPrompt that guides the video generation process via optical flow. Specifically, we train a discriminator to distinguish optical flow between random pairs of frames from real videos and generated ones. Given that prompts can influence the entire video, we optimize learnable token embeddings during reverse sampling steps by using gradients from a trained discriminator applied to random frame pairs. This approach allows our method to generate visually coherent video sequences that closely reflect natural motion dynamics, without compromising the fidelity of the generated content. We demonstrate the effectiveness of our approach across various models.

cs.CV↗

arXiv2025

Purification of Noisy Measurements and Faithful Distillation of Entanglement

We consider entanglement distillation with noisy operations in which quantum measurements that constitute a general quantum operation are particularly noisy. We present a protocol for purifying noisy measurements and show that imperfect local operations can distill entanglement. The protocol works for arbitrary noisy measurements in general and is cost-effective and resource-efficient with single additional qubit per party to resolve the distillation of entanglement. The purification protocol is feasible with currently available quantum technologies and readily applied to entanglement applications.

quant-ph↗

arXiv2024

Static Quantum Errors and Purification

State preparation that initializes quantum systems in a fiducial state and measurements to read outcomes after the evolution of quantum states, both essential elements in quantum information processing in general, may contain noise from which errors, in particular, referred to as static errors, may appear even with noise-free evolution. In this work, we consider noisy resources such as faulty manufacturing and improper maintenance of systems by which static errors in state preparation and measurement are inevitable. We show how to suppress static errors and purify noiseless SPAM by repeatedly applying noisy resources. We present the purification protocol for noisy initialization and noisy measurements and verify that a few qubits are immediately cost-effective to suppress error rates up to $10^{-3}$. We also demonstrate the purification protocol in a realistic scenario. The results are readily feasible with current quantum technologies.

quant-ph↗

arXiv2024

SentiCSE: A Sentiment-aware Contrastive Sentence Embedding Framework with Sentiment-guided Textual Similarity

Recently, sentiment-aware pre-trained language models (PLMs) demonstrate impressive results in downstream sentiment analysis tasks. However, they neglect to evaluate the quality of their constructed sentiment representations; they just focus on improving the fine-tuning performance, which overshadows the representation quality. We argue that without guaranteeing the representation quality, their downstream performance can be highly dependent on the supervision of the fine-tuning data rather than representation quality. This problem would make them difficult to foray into other sentiment-related domains, especially where labeled data is scarce. We first propose Sentiment-guided Textual Similarity (SgTS), a novel metric for evaluating the quality of sentiment representations, which is designed based on the degree of equivalence in sentiment polarity between two sentences. We then propose SentiCSE, a novel Sentiment-aware Contrastive Sentence Embedding framework for constructing sentiment representations via combined word-level and sentence-level objectives, whose quality is guaranteed by SgTS. Qualitative and quantitative comparison with the previous sentiment-aware PLMs shows the superiority of our work. Our code is available at: https://github.com/nayohan/SentiCSE

cs.CL↗

arXiv2024

ExeGPT: Constraint-Aware Resource Scheduling for LLM Inference

This paper presents ExeGPT, a distributed system designed for constraint-aware LLM inference. ExeGPT finds and runs with an optimal execution schedule to maximize inference throughput while satisfying a given latency constraint. By leveraging the distribution of input and output sequences, it effectively allocates resources and determines optimal execution configurations, including batch sizes and partial tensor parallelism. We also introduce two scheduling strategies based on Round-Robin Allocation and Workload-Aware Allocation policies, suitable for different NLP workloads. We evaluate ExeGPT on six LLM instances of T5, OPT, and GPT-3 and five NLP tasks, each with four distinct latency constraints. Compared to FasterTransformer, ExeGPT achieves up to 15.2x improvements in throughput and 6x improvements in latency. Overall, ExeGPT achieves an average throughput gain of 2.9x across twenty evaluation scenarios. Moreover, when adapting to changing sequence distributions, the cost of adjusting the schedule in ExeGPT is reasonably modest. ExeGPT proves to be an effective solution for optimizing and executing LLM inference for diverse NLP workload and serving conditions.

cs.DC↗

arXiv2023

Modeling the mechanosensitive collective migration of cells on the surface and the interior of morphing soft tissues

Cellular contractility, migration, and extracellular matrix (ECM) mechanics are critical for a wide range of biological processes including embryonic development, wound healing, tissue morphogenesis, and regeneration. Even though the distinct response of cells near the tissue periphery has been previously observed in cell-laden microtissues, including faster kinetics and more prominent cell-ECM interactions, there are currently no models that can fully combine coupled surface and bulk mechanics and kinetics to recapitulate the morphogenic response of these constructs. Mailand \textit{et al.} (2019) had shown the importance of active elastocapillarity in cell-laden microtissues, but modeling the distinct mechanosensitive migration of cells on the periphery and the interior of highly deforming tissues has not been possible thus fur, especially in the presence of active elastocapillary effects. This paper presents a framework for understanding the interplay between cellular contractility, migration, and ECM mechanics in dynamically morphing soft tissues accounting for distinct cellular responses in the bulk and the surface of tissues. The major novelty of this approach is that it enables modeling the distinct migratory and contractile response of cells residing on the tissue surface and the bulk, where concurrently the morphing soft tissues undergoes large deformations driven by cell contractility. Additionally, the proposed model is validated through simulation results that capture the changes in shape and cell concentration for wounded and intact microtissues, enabling the interpretation of experimental data.

physics.bio-ph↗