Searcharxiv⌕ Search

arXiv subjects

Minsu Kim

Publications and source records attributed to Minsu Kim.

At least 19 recordsLinked to original sources
arXiv2026

Preference Data Selection for Mitigating the Alignment Tax in Large Language Models

Aligning large language models to human preferences is crucial for real-world deployment but frequently incurs an alignment tax, leading to the catastrophic forgetting of pre-trained general capabilities. While previous works primarily frame this problem as an optimization or architectural challenge, the inherent characteristics of preference data that drive this degradation remain largely underexplored. In this paper, we propose BALIGN, a balanced data selection strategy that explicitly mitigates catastrophic forgetting while optimizing alignment efficacy. Through theoretical and empirical analyses of the preference optimization gradient, we identify three key data-centric features that dictate parameter drift: the reference model's log-probability margin, the token length difference between chosen and rejected responses, and the TF-IDF similarity to general capability corpora. By aggregating these orthogonal features into a unified composite risk score, BALIGN systematically filters out high-risk preference samples that disrupt intrinsic model parameters or provide minimal alignment utility. Extensive experiments on standard human preference datasets demonstrate that BALIGN strongly preserves foundational capabilities without compromising alignment gains, consistently achieving the optimal Pareto frontier with minimal computational overhead.

cs.AI↗
arXiv2026

Simthesizer: An Agent-Driven Simulation Framework for LLM Serving Systems

System-level simulation is an essential tool for exploring the rapidly expanding design space of LLM serving systems, where real deployments remain costly and often infeasible. However, modern LLM serving now evolves faster than human-driven simulator development can track, and emerging workloads and mechanisms, from agentic workflows to disaggregated serving, no longer fit the monolithic simulation pipeline that existing simulators assume. Each new mechanism therefore demands an invasive rewrite, leaving a widening development gap between deployed serving systems and the simulators that model them. To close this gap, we present Simthesizer, a framework that realizes agent-driven simulator development. Simthesizer introduces a composable simulator infrastructure that uniformly expresses the complete serving workflow, including the control decisions that coordinate it, and realizes it as a unified dynamic graph in Simthesizer simulator. Synthesizer agent, a harnessed coding agent, then lowers natural-language feature requests onto this abstraction under simulator-specific guardrails and fidelity validation, evolving one shared simulator instead of building a new one for every feature. Under the same coding agent and harnesses, extensions built on Simthesizer follow a vLLM-based real system with 2.51% average throughput error, versus 6.03% for extensions built on existing simulators. On identical workloads, Simthesizer also simulates up to 284.96x and 23.19x faster than two state-of-the-art simulators, LLMServingSim2.0 and Vidur, respectively.

cs.AR↗
arXiv2026

Exact No Signaling in Time without Temporal Classicality

No signaling in time (NSIT) has often been treated as the clean operational remnant of noninvasive measurability. If an earlier measurement leaves every later marginal unchanged, the temporal process appears classical. We show that this inference is false. We introduce common fixed points (CFPs) of nonselective measurement channels as an exact mechanism that erases all marginal evidence of invasiveness while preserving disturbed outcome conditioned branches. For a qutrit ring subject to a local degenerate Luders measurement, we solve the full CFP manifold analytically. Every state on this manifold satisfies exact pairwise NSIT, yet every nontrivial member violates a Leggett Garg inequality, including the maximally mixed state. The violation is governed by finite branch displacement, not by residual signaling. Hidden variable reconstruction, entropic witnesses, protocol landscape scans, noise robustness, and finite shot simulations show that exact NSIT certifies only the disappearance of marginal signals after outcome erasure, not the existence of a classical temporal history.

quant-ph↗
arXiv2026

Process-Verified Reinforcement Learning for Theorem Proving via Lean

While reinforcement learning from verifiable rewards (RLVR) typically has relied on a single binary verification signal, symbolic proof assistants in formal reasoning offer rich, fine-grained structured feedback. This gap between structured processes and unstructured rewards highlights the importance of feedback that is both dense and sound. In this work, we demonstrate that the Lean proof assistant itself can serve as a symbolic process oracle, supplying both outcome-level and fine-grained tactic-level verified feedback during training. Proof attempts are parsed into tactic sequences, and Lean's elaboration marks both locally sound steps and the earliest failing step, yielding dense, verifier-grounded credit signals rooted in type theory. We incorporate these structured rewards into a GRPO-style reinforcement learning objective with first-error propagation and first-token credit methods that balances outcome- and process-level advantages. Experiments with STP-Lean and DeepSeek-Prover-V1.5 show that tactic-level supervision outperforms outcome-only baselines in most settings, delivering improvements on benchmarks such as MiniF2F and ProofNet. Beyond empirical gains, our study highlights a broader perspective: symbolic proof assistants are not only verifiers at evaluation time, but can also act as process-level reward oracles during training. This opens a path toward reinforcement learning frameworks that combine the scalability of language models with the reliability of symbolic verification for formal reasoning.

cs.AI↗
arXiv2026

Decoding Strategies for Diffusion-Based ASR: A Systematic Evaluation of Confidence-Based Thresholding

While LLM-based Automatic Speech Recognition (ASR) achieves high accuracy, its speed is limited by sequential autoregressive decoding. Diffusion Language Models (DLMs) offer a parallel alternative, yet their decoding strategies remain under-explored in ASR contexts. This paper analyzes three decoding schemes for DLM-based ASR: fixed-number, static confidence threshold, and dynamic confidence threshold. We introduce a round-wise analysis of decoding progress using Negative Log-Likelihood-based uncertainty as a proxy for prediction reliability. Our results show that both threshold-based strategies provide a better accuracy-speed trade-off than fixed-number schemes. This behavior is associated with the more concentrated confidence distribution observed in the evaluated ASR settings: many tokens reach high confidence early, enabling multiple tokens to be committed in early decoding rounds while lower-confidence tokens are deferred to later rounds. The static-threshold strategy achieves accuracy close to autoregressive decoding at lower decoding cost.

eess.AS↗
arXiv2026

modelimportance: An R package for evaluating model importance within a multi-model ensemble

Ensemble forecasts are commonly used to support decision-making and policy planning across various fields because they often offer improved accuracy and stability compared to individual models. As each model has its own unique characteristics, understanding and measuring the value of each constituent model can support the construction of effective ensembles. The R package modelimportance provides tools to quantify how each component model contributes to the accuracy of ensemble performance for both point and probabilistic forecasts. The package supports multiple ensemble methods and multiple model importance metrics. Additionally, the software offers customizable options for handling missing values. These features enable the package to serve as a versatile tool for researchers and practitioners. It helps not only in constructing an effective ensemble model across a wide range of forecasting tasks, but also in understanding the role of each model within the ensemble and gaining insights into individual models themselves. This package follows the 'hubverse' framework, which is a collection of open-source software, tools and data standards developed to promote collaborative modeling hub efforts and simplify their setup and operation. Doing so enables seamless integration and flexibility with other forecasting tools and systems, allowing many analyses to be performed on existing hubs.

stat.CO↗
arXiv2026

Aligning Few-Step Generative Models by Amortizing Sample-based Variational Inference

Aligning a few-step generative model is challenging, since existing alignment frameworks typically rely on restrictive assumptions: a tractable likelihood, a specific ODE/SDE solver, or a particular model family. We introduce FAV, Few-step Generative Models Alignment via Sample-based Variational Inference, a general alignment framework that requires only sample access to the generator and the reference distribution. We cast alignment as sampling from a reward-tilted distribution anchored to a reference distribution. We leverage Stein Variational Gradient Descent as a sample-based variational inference scheme and amortize its particle updates into the generator parameters via fixed-point regression. We evaluate FAV on two domains: robotics manipulation and image generator alignment. On generative policy alignment for robotic manipulation, FAV outperforms prevailing policy extraction baselines across 56 offline and 30 offline-to-online RL tasks. For image generator alignment, FAV fine-tunes diverse few-step backbones, including GAN, drifting model, consistency models, and flow maps, scaling from ImageNet-$256$ to 1024$^2$ text-to-image synthesis. Code is available at https://github.com/Jaewoopudding/FAV.

cs.LG↗
arXiv2026

Generative Recursive Reasoning

How should future neural reasoning systems implement extended computation? Recursive Reasoning Models (RRMs) offer a promising alternative to autoregressive sequence extension by performing iterative latent-state refinement with shared transition functions. Yet existing RRMs are largely deterministic, following a single latent trajectory and converging to a single prediction. We introduce Generative Recursive reAsoning Models (GRAM), a framework that turns recursive latent reasoning into probabilistic multi-trajectory computation. GRAM models reasoning as a stochastic latent trajectory, enabling multiple hypotheses, alternative solution strategies, and inference-time scaling through both recursive depth and parallel trajectory sampling. This yields a latent-variable generative model supporting conditional reasoning via $p_\theta(y \mid x)$ and, with fixed or absent inputs, unconditional generation via $p_\theta(x)$. Trained with amortized variational inference, GRAM improves over deterministic recurrent and recursive baselines on structured reasoning and multi-solution constraint satisfaction tasks, while demonstrating an unconditional generation capability. https://ahn-ml.github.io/gram-website

cs.AI↗
arXiv2026

Transformer Architecture with Minimal Inference Latency for Multi-Modal Wireless Networks

Next-generation wireless networks are expected to leverage multi-modal data sources to execute various wireless communication tasks such as beamforming and blockage prediction with situational-awareness. To do so, multi-modal transformers emerged as an effective tool, however, existing transformer-based approaches suffer from high inference latency and large memory footprints when processing multi-modal data. Hence, such existing solutions cannot handle wireless communication tasks that require fast inference to track a dynamically changing environment with moving vehicles and blockages. One major bottleneck is the reliance on attention mechanisms whose complexity grows quadratically with respect to the number of tokens. Hence, in this paper, a novel, fast multi-modal transformer inference framework is designed to practically support wireless communication tasks by processing only important tokens. To this end, an optimization problem is formulated to find the optimal number of tokens under a target FLOPs for a given wireless communication task while maintaining the task accuracy. To solve this problem, modality-specific tokenizers are first designed to project each modality into the same embedding dimension. Then, a token router is introduced to learn the importance of each token and process only important tokens. Subsequently, a trainable keep ratio is introduced to learn how many tokens to process for each layer under the target FLOPs. Simulation results show that, on DeepSense 6G beamforming tasks, we can reduce the inference latency, GPU memory, and FLOPs by 86.2% 35%, and 80%, respectively, with negligible accuracy loss. To validate the feasibility for real-world deployments, a multi-modal handover dataset is developed using a real-world testbed. Emulation results on the developed dataset show that the proposed framework can proactively initiate handover before blockage.

eess.SY↗
arXiv2026

UniSpector: Towards Universal Open-set Defect Recognition via Spectral-Contrastive Visual Prompting

Although industrial inspection systems should be capable of recognizing unprecedented defects, most existing approaches operate under a closed-set assumption, which prevents them from detecting novel anomalies. While visual prompting offers a scalable alternative for industrial inspection, existing methods often suffer from prompt embedding collapse due to high intra-class variance and subtle inter-class differences. To resolve this, we propose UniSpector, which shifts the focus from naive prompt-to-region matching to the principled design of a semantically structured and transferable prompt topology. UniSpector employs the Spatial-Spectral Prompt Encoder to extract orientation-invariant, fine-grained representations; these serve as a solid basis for the Contrastive Prompt Encoder to explicitly regularize the prompt space into a semantically organized angular manifold. Additionally, Prompt-guided Query Selection generates adaptive object queries aligned with the prompt. We introduce Inspect Anything, the first benchmark for visual-prompt-based open-set defect localization, where UniSpector significantly outperforms baselines by at least 19.7% and 15.8% in AP50b and AP50m, respectively. These results show that our method enable a scalable, retraining-free inspection paradigm for continuously evolving industrial environments, while offering critical insights into the design of generic visual prompting.

cs.CV↗
arXiv2026

STRIDE: When to Speak Meets Sequence Denoising for Streaming Video Understanding

Recent progress in video large language models (Video-LLMs) has enabled strong offline reasoning over long and complex videos. However, real-world deployments increasingly require streaming perception and proactive interaction, where video frames arrive online and the system must decide not only what to respond, but also when to respond. In this work, we revisit proactive activation in streaming video as a structured sequence modeling problem, motivated by the observation that temporal transitions in streaming video naturally form span-structured activation patterns. To capture this span-level structure, we model activation signals jointly over a sliding temporal window and update them iteratively as new frames arrive. We propose STRIDE (Structured Temporal Refinement with Iterative DEnoising), which employs a lightweight masked diffusion module at the activation interface to jointly predict and progressively refine activation signals across the window. Extensive experiments on diverse streaming benchmarks and downstream models demonstrate that STRIDE shows more reliable and temporally coherent proactive responses, significantly improving when-to-speak decision quality in online streaming scenarios.

cs.CV↗
arXiv2026

Na\"ive Exposure of Generative AI Capabilities Undermines Deepfake Detection

Generative AI systems increasingly expose powerful reasoning and image refinement capabilities through user-facing chatbot interfaces. In this work, we show that the na\"ive exposure of such capabilities fundamentally undermines modern deepfake detectors. Rather than proposing a new image manipulation technique, we study a realistic and already-deployed usage scenario in which an adversary uses only benign, policy-compliant prompts and commercial generative AI systems. We demonstrate that state-of-the-art deepfake detection methods fail under semantic-preserving image refinement. Specifically, we show that generative AI systems articulate explicit authenticity criteria and inadvertently externalize them through unrestricted reasoning, enabling their direct reuse as refinement objectives. As a result, refined images simultaneously evade detection, preserve identity as verified by commercial face recognition APIs, and exhibit substantially higher perceptual quality. Importantly, we find that widely accessible commercial chatbot services pose a significantly greater security risk than open-source models, as their superior realism, semantic controllability, and low-barrier interfaces enable effective evasion by non-expert users. Our findings reveal a structural mismatch between the threat models assumed by current detection frameworks and the actual capabilities of real-world generative AI. While detection baselines are largely shaped by prior benchmarks, deployed systems expose unrestricted authenticity reasoning and refinement despite stringent safety controls in other domains.

cs.CR↗
arXiv2026

IMSE: Intrinsic Mixture of Spectral Experts Fine-tuning for Test-Time Adaptation

Test-time adaptation (TTA) has been widely explored to prevent performance degradation when test data differ from the training distribution. However, fully leveraging the rich representations of large pretrained models with minimal parameter updates remains underexplored. In this paper, we propose Intrinsic Mixture of Spectral Experts (IMSE) that leverages the spectral experts inherently embedded in Vision Transformers. We decompose each linear layer via singular value decomposition (SVD) and adapt only the singular values, while keeping the singular vectors fixed. We further identify a key limitation of entropy minimization in TTA: it often induces feature collapse, causing the model to rely on domain-specific features rather than class-discriminative features. To address this, we propose a diversity maximization loss based on expert-input alignment, which encourages diverse utilization of spectral experts during adaptation. In the continual test-time adaptation (CTTA) scenario, beyond preserving pretrained knowledge, it is crucial to retain and reuse knowledge from previously observed domains. We introduce Domain-Aware Spectral Code Retrieval, which estimates input distributions to detect domain shifts, and retrieves adapted singular values for rapid adaptation. Consequently, our method achieves state-of-the-art performance on various distribution-shift benchmarks under the TTA setting. In CTTA and Gradual CTTA, it further improves accuracy by 3.4 percentage points (pp) and 2.4 pp, respectively, while requiring 385 times fewer trainable parameters. Our code is available at https://github.com/baek85/IMSE.

cs.CV↗
arXiv2026

Optimal trajectory-guided stochastic co-optimization for e-fuel system design and real-time operation

E-fuels are promising long-term energy carriers supporting the net-zero transition. However, the large combinatorial design-operation spaces under renewable uncertainty make the use of mathematical programming impractical for co-optimizing e-fuel production systems. Here, we present MasCOR, a machine-learning-assisted co-optimization framework that learns from global operational trajectories. By encoding system design and renewable trends, a single MasCOR agent generalizes dynamic operation across diverse configurations and scenarios, substantially simplifying design-operation co-optimization under uncertainty. Benchmark comparisons against state-of-the-art reinforcement learning baselines demonstrate near-optimal performance, while computational costs are substantially lower than those of mathematical programming, enabling rapid parallel evaluation of designs within the co-optimization loop. This framework enables rapid screening of feasible design spaces together with corresponding operational policies. When applied to four potential European sites targeting e-methanol production, MasCOR shows that most locations benefit from reducing system load below 50 MW to achieve carbon-neutral methanol production, with production costs of 1.0-1.2 USD per kg. In contrast, Dunkirk (France), with limited renewable availability and high grid prices, favors system loads above 200 MW and expanded storage to exploit dynamic grid exchange and hydrogen sales to the market. These results underscore the value of the MasCOR framework for site-specific guidance from system design to real-time operation.

cs.LG↗
arXiv2026

AVID: A Near-Major Post-Merger of Late-Type Dwarfs beneath a Regularly Rotating HI Disk (VCC 693)

On the periphery of galaxy clusters, moderately high galaxy densities and velocity dispersions favour interactions and mergers that influence galaxy evolution prior to cluster infall. Observational studies of this phase in dwarfs remain rare. We present a high-resolution study of the merger remnant VCC 693 in the outskirts of Virgo cluster, using observations from the Atomic gas in Virgo Interacting Dwarf galaxies (AVID) project. We explore the origin of VCC 693 and the consequences of the merger on its star formation and structure through a joint analysis of VLA and FAST HI emission line observations, together with complementary optical imaging and spectroscopy. We employ hydrodynamical simulations to help interpret the observations. Our analysis favours a near-major merger between two dwarfs with a stellar mass ratio of 3:1-4:1, with one likely gas-poor progenitor (i.e., a damp merger). The optical appearance of VCC 693 is dominated by complex tidal structures throughout the system, whereas the HI gas has settled to a regular rotating disk. Compared with similar-mass dwarfs, the central star formation and gas-phase metallicity are moderately enhanced. The global star formation rate, HI gas content, and HI-to-optical size ratio of VCC 693 are broadly consistent with those of typical dwarfs of similar mass, albeit somewhat lower. Decomposition of the HI rotation curve into baryonic and dark matter indicates a high halo concentration, suggesting post-merger relaxation into a more centrally peaked configuration. Together with two recent studies of AVID post-merger systems, these results support the view that even major dwarf mergers can produce remnants with overall stellar structures indistinguishable from ordinary dwarfs, and that the environmental effects in cluster outskirts can promote damp or mixed mergers, constituting an integral part of galactic pre-processing.

astro-ph.GA↗
arXiv2026

Synthesizable Molecular Generation via Soft-constrained GFlowNets with Rich Chemical Priors

The application of generative models for experimental drug discovery campaigns is severely limited by the difficulty of designing molecules de novo that can be synthesized in practice. Previous works have leveraged Generative Flow Networks (GFlowNets) to impose hard synthesizability constraints through the design of state and action spaces based on predefined reaction templates and building blocks. Despite the promising prospects of this approach, it currently lacks flexibility and scalability. As an alternative, we propose S3-GFN, which generates synthesizable SMILES molecules via simple soft regularization of a sequence-based GFlowNet. Our approach leverages rich molecular priors learned from large-scale SMILES corpora to steer molecular generation towards high-reward, synthesizable chemical spaces. The model induces constraints through off-policy replay training with a contrastive learning signal based on separate buffers of synthesizable and unsynthesizable samples. Our experiments show that S3-GFN learns to generate synthesizable molecules ($\geq 95\%$) with higher rewards in diverse tasks.

cs.LG↗
arXiv2026

A generalizable large-scale foundation model for musculoskeletal radiographs

Artificial intelligence (AI) has shown promise in detecting and characterizing musculoskeletal diseases from radiographs. However, most existing models remain task-specific, annotation-dependent, and limited in generalizability across diseases and anatomical regions. Although a generalizable foundation model trained on large-scale musculoskeletal radiographs is clinically needed, publicly available datasets remain limited in size and lack sufficient diversity to enable training across a wide range of musculoskeletal conditions and anatomical sites. Here, we present SKELEX, a large-scale foundation model for musculoskeletal radiographs, trained using self-supervised learning on 1.2 million diverse, condition-rich images. The model was evaluated on 12 downstream diagnostic tasks and generally outperformed baselines in fracture detection, osteoarthritis grading, and bone tumor classification. Furthermore, SKELEX demonstrated zero-shot abnormality localization, producing error maps that identified pathologic regions without task-specific training. Building on this capability, we developed an interpretable, region-guided model for predicting bone tumors, which maintained robust performance on independent external datasets and was deployed as a publicly accessible web application. Overall, SKELEX provides a scalable, label-efficient, and generalizable AI framework for musculoskeletal imaging, establishing a foundation for both clinical translation and data-efficient research in musculoskeletal radiology.

cs.CV↗
arXiv2026

GFlowPO: Generative Flow Network as a Language Model Prompt Optimizer

Finding effective prompts for language models (LMs) is critical yet notoriously difficult: the prompt space is combinatorially large, rewards are sparse due to expensive target-LM evaluation. Yet, existing RL-based prompt optimizers often rely on on-policy updates and a meta-prompt sampled from a fixed distribution, leading to poor sample efficiency. We propose GFlowPO, a probabilistic prompt optimization framework that casts prompt search as a posterior inference problem over latent prompts regularized by a meta-prompted reference-LM prior. In the first step, we fine-tune a lightweight prompt-LM with an off-policy Generative Flow Network (GFlowNet) objective, using a replay-based training policy that reuses past prompt evaluations to enable sample-efficient exploration. In the second step, we introduce Dynamic Memory Update (DMU), a training-free mechanism that updates the meta-prompt by injecting both (i) diverse prompts from a replay buffer and (ii) top-performing prompts from a small priority queue, thereby progressively concentrating the search process on high-reward regions. Across few-shot text classification, instruction induction benchmarks, and question answering tasks, GFlowPO consistently outperforms recent discrete prompt optimization baselines.

cs.AI↗