Searcharxiv⌕ Search

arXiv subjects

Ke Zhang

Publications and source records attributed to Ke Zhang.

At least 55 records · Page 3Linked to original sources
arXiv2026

From Young to Older Disks: JWST/MIRI Evidence for Fading Molecular Emission and Hints for Elevated C/O in Upper Scorpius

We present JWST/MIRI spectroscopy of 14 disks in the older (~5-10 Myr) Upper Scorpius (USco) association and use slab of gas in local thermal equilibrium to infer basic gas properties. We find that half of these disks are molecular rich, with detections of H$_2$O, CO$_2$, HCN, C$_2$H$_2$, and H$_2$, while the other half are molecular poor, showing no molecular emission other than H$_2$. We further combine this sample with 10 other USco disks from the AGE-PRO program and compare the combined older sample to young (~1-3 Myr) JDISCS Cycle~1 systems, which are analyzed in a similar manner. We find that USco disks have lower detection rates of major molecular species but a significantly higher detection rate of rarer C-bearing molecules such as C$_4$H$_2$. At a given accretion luminosity, molecular line luminosities are systematically lower in USco than in young disks, and the scaling relations with accretion luminosity differ between the two populations. Moreover, we find that about half of the older disks, preferentially the millimeter faint, and likely more compact disks, have observable mass ratios of C- to O-bearing molecules that are higher than the maximum values in the young sample. These results point to reduced inner-disk molecular gas masses, cooler emitting layers, and higher inner gas C/O ratios in older disks, the latter being consistent with pebble drift. Taken together, our findings provide evidence for chemical evolution of inner disk gas from young to older systems, with important implications for the accretion of primordial planetary atmospheres.

astro-ph.EP↗
arXiv2026

Molecular Similarity and Water Diversity in Coeval Binary Disks: JWST/MIRI Observations of Sz 65 and Sz 66

We present JWST/MIRI Medium Resolution Spectrometer spectra of the wide-separation (projected separation $= 980$ au) binary protoplanetary disks Sz 65 (K7; $0.68~M_{\odot}$) and Sz 66 (M3; $0.30~M_{\odot}$), reduced using the uniform pipeline of the JWST Disk Infrared Spectral Chemistry Survey. Both disks show rich molecular emission, including H$_2$O, CO$_2$, HCN, C$_2$H$_2$, and OH. The scaled spectra of the two disks exhibit remarkably similar H$_2$O, CO$_2$, and HCN line emission in the 13--18 $μ$m region, with the only notable difference being stronger C$_2$H$_2$ emission in the primary (Sz 65). Beyond 18 $μ$m, the difference in H$_2$O line emission between the two disks increases. Both the flux ratios and the slab-model-derived mass ratios of cold to hot H$_2$O ($\sim$200 K to $\sim$750 K) and warm to hot H$_2$O ($\sim$450 K to $\sim$750 K) are significantly higher in the secondary (Sz 66). Because binary stars share nearly the same age and metallicity, and as both disks appear compact in millimeter emission ($<30$ au), we suggest that the excess cold H$_2$O in the secondary is best explained by its unstructured dust disk, in contrast to the primary, which shows gaps at 6 and 20 au. The enhanced cold water in the secondary is consistent with efficient pebble drift across the water snow line and increased H$_2$O vapor from the sublimation of icy mantles. Our results demonstrate that wide-separation binaries can serve as powerful control samples for isolating the impact of individual disk properties on inner-disk chemistry and evolution.

astro-ph.EP↗
arXiv2026

FlowTrain: Flow-Based Decoupled Training for Industrial-Grade Vision-Language Models

Industrial-grade distributed training of vision-language models (VLMs) remains far less efficient than that of unimodal LLMs. Existing solutions either follow a monolithic design that assigns uniform parallelism to heterogeneous modules or adopt a disaggregated deployment that separates modules while executing them as a batch-synchronized pipeline. In this paper, we highlight that the above solutions are still not sufficient, and VLM training can be further decoupled. To this end, we present FlowTrain, a flow-based decoupled training framework that reformulates VLM training as a producer-consumer dataflow coordinated through a unified memory pool. The encoder and backbone can progress independently over a global virtual address space. Since this execution decoupling fundamentally changes the optimization objective of allocation and scheduling, FlowTrain further introduces a heterogeneous parallel allocator that assigns module-specific parallelism strategies by solving a throughput matching problem. The dynamic packing scheduler is used to construct balanced microbatches at runtime according to the actual LLM-side computation cost. Extensive experiments on real-world workloads show that FlowTrain achieves over 50% MFU and up to 1.7x throughput improvement, narrowing the efficiency gap to LLM-only training.

cs.LG↗
arXiv2026

BLADE: Scalable Bi-level Adaptive Data Selection for LLM Training

As Large Language Model (LLM) datasets scale to trillions of tokens, data selection has emerged as a critical frontier to filter out uninformative noise and construct adaptive learning trajectories. Beyond static heuristic filtering, advanced data selection methods for LLM training largely follow two paradigms, each with fundamental limitations. Influence-based methods provide principled bi-level objectives but require intractable inverse-Hessian computations, while excess-loss methods are computationally efficient but rely on a static reference model that becomes misaligned with the evolving proxy model during training. We propose BLADE (Bi-Level Adaptive Data sElection), a Hessian-free framework for data selection. BLADE reformulates the bi-level optimization problem underlying influence-based methods as a penalized single-level objective via Lagrange multipliers, avoiding inverse-Hessian computation while revealing a principled connection to excess-loss based data selection. The resulting objective recovers an excess-loss form but replaces the static reference model with a dynamic one that stays synchronized with training. Theoretically, we prove that this penalized formulation guarantees first-order convergence. For efficient online batch selection, we instantiate BLADE as a memoryless randomized block-coordinate Frank-Wolfe algorithm. Extensive experiments show that BLADE consistently outperforms state-of-the-art data selection baselines, providing a practical recipe for LLM training.

cs.LG↗
arXiv2026

3D Consistency Optimization for Self-Supervised Monocular Video Depth Estimation

Reliable monocular video depth estimation is crucial for downstream 3D reasoning and embodied AI in endoscopic navigation. However, existing self-supervised approaches typically treat video frames independently or rely on weak temporal regularization. These methods, lacking a holistic perception of the underlying 3D scene, inevitably suffer from geometrically inconsistent predictions and severe cross-frame drift. To address these limitations, we introduce a new paradigm that recasts sequential video depth estimation as an unconstrained multi-view 3D reconstruction problem, enabling full exploitation of the powerful geometric priors embedded in recent 3D foundation models. The core of our approach is a 3D consistency optimization framework driven by three constraints: image-level photometric rendering, explicit world-coordinate geometric alignment, and multi-scale temporal gradient consistency. Such unified optimization elegantly anchors isolated frames to a globally coherent 3D structure. Our method has been validated in both the self-supervised training scenarios and challenging zero-shot clinical environments. Results show that the proposed approach achieves state-of-the-art spatial accuracy, outperforming the frame-based, video-based depth estimators and the multi-view 3D reconstruction baselines.

cs.CV↗
arXiv2026

Multi-channel high-speed flip-chip packaging platform for thin-film lithium niobate photonic circuits

To address the urgent need for multi-channel high-speed electrical interfacing of thin-film lithium niobate (TFLN) photonic circuits, we realize a flip-chip packaging platform capable of simultaneously delivering 13 high-speed and 32 low-speed electronic signals to a centimeter-sized TFLN chip. The platform exhibits low flip-chip bonding loss and low inter-channel crosstalk over a broad bandwidth up to 50 GHz. Leveraging this packaging platform, we demonstrate high-speed electrical interfacing with two proof-of-concept TFLN photonic circuits, namely a 2x8 optical switch and an electro-optic comb-based transmitter. The switch achieves arbitrary 8-channel routing with ~3 dB insertion loss, < -20 dB crosstalk, and an equipment-limited switching time of <= 34 ps. The transmitter circuit includes a 50 GHz electro-optic comb generator with 2.8-dB flatness, a tunable microring to arbitrarily filter one comb line, and a modulator for data transmission at 20 Gbit/s. The packaging platform could significantly advance large-scale TFLN circuits in optical communications, microwave photonics, and photonic computing.

physics.optics↗
arXiv2026

ProServe: Unified Multi-Priority Request Scheduling for LLM Serving

The widespread deployment of large language models (LLMs) for interactive applications necessitates serving systems that can handle thousands of concurrent requests with diverse Service Level Objective (SLO) requirements. A critical yet often overlooked dimension in this context is the inherent priority difference among clients; for instance, business-critical functions demand higher performance guarantees, as fulfilling such requests yields significantly greater business value. However, existing LLM serving schedulers fail to jointly optimize for both SLO attainment and client-level priorities. To bridge this gap, we first formalize multi-priority request scheduling as a service gain maximization problem, where satisfying latency requirements for requests of different priorities contributes varying gain. We propose ProServe, a unified two-tier scheduling framework designed to maximize overall service gain. At the engine layer, SlideBatching dynamically adapts batch formation under varying loads, employing a sliding boundary mechanism to balance latency and priority differentiation. Considering potential preemption, efficient block management adopts asynchronous offloading, pipelined reloading, and adaptive copy-budget control to overlap computation with host-device block transfers. At the service layer, GoRouting performs gain-oriented and capability-aware dispatching across distributed instances, proactively reserving capacity for future high-priority or long requests. Extensive evaluation on four open-source and one industrial dataset shows that ProServe outperforms state-of-the-art baselines, improving system gain by up to 35% and SLO attainment by up to 52%.

cs.DC↗
arXiv2026

TANDEM: Bi-Level Data Mixture Optimization with Twin Networks

The capabilities of large language models (LLMs) significantly depend on training data drawn from various domains. Optimizing domain-specific mixture ratios can be modeled as a bi-level optimization problem, which we simplify into a single-level penalized form and solve with twin networks: a proxy model trained on primary data and a dynamically updated reference model trained with additional data. Our proposed method, Twin Networks for bi-level DatA mixturE optiMization (TANDEM), measures the data efficacy through the difference between the twin models and up-weights domains that benefit more from the additional data. TANDEM provides theoretical guarantees and wider applicability, compared to prior approaches. Furthermore, our bi-level perspective suggests new settings to study domain reweighting such as data-restricted scenarios and supervised fine-tuning, where optimized mixture ratios significantly improve the performance. Extensive experiments validate TANDEM's effectiveness in all scenarios.

cs.LG↗
arXiv2026

The ALMA Survey of Gas Evolution of PROtoplanetary Disks (AGE-PRO): Formaldehyde (H$_2$CO) emission and its links to disk properties

Protoplanetary disks are rotating structures of gas and dust surrounding young stars, serving as the birth places of planets. Understanding the chemical evolution of organic materials in these disks is key for tracing the origins of organics in planetary systems. Formaldehyde (H$_2$CO) is the most commonly detected organic molecule in protoplanetary disks. In this study, we investigate the emission of H$_2$CO and its link to disk properties, using a sample of 20 Class II disks in the Lupus and Upper Sco star-forming regions spanning over 1-6 Myr. We analyze the H$_2$CO lines at 218.222 and 290.623 GHz observed as part of the AGE-PRO ALMA Large Program. Within this sample we achieve a detection rate of H$_2$CO of 45% (9/20), and set robust upper limits for the non-detections. We measure the excitation temperature and column density of the H$_2$CO gas in the sources with H$_2$CO detections. We combine our sample with 13 additional disks with archival H$_2$CO detections and search for correlations between H$_2$CO properties and disk parameters. Notably, we find strong correlations between H$_2$CO line luminosity and dust radius, gas radius, dust mass, gas mass, stellar mass, and stellar luminosity. This suggests that H$_2$CO emission is brighter for extended massive dust disks where H$_2$CO can form via CO ice hydrogenation on grain surfaces. We find that the H$_2$CO excitation temperature is also correlated with stellar mass and stellar luminosity, so more massive and luminous stars could increase H$_2$CO excitation.

astro-ph.EP↗
arXiv2026

Revisiting Model Stitching In the Foundation Model Era

Model stitching, connecting early layers of one model (source) to later layers of another (target) via a light stitch layer, has served as a probe of representational compatibility. Prior work finds that models trained on the same dataset remain stitchable (negligible accuracy drop) despite different initializations or objectives. We revisit stitching for Vision Foundation Models (VFMs) that vary in objectives, data, and modality mix (e.g., CLIP, DINOv2, SigLIP 2) and ask: Are heterogeneous VFMs stitchable? We introduce a systematic protocol spanning the stitch points, stitch layer families, training losses, and downstream tasks. Three findings emerge. (1) Stitch layer training matters: conventional approaches that match the intermediate features at the stitch point or optimize the task loss end-to-end struggle to retain accuracy, especially at shallow stitch points. (2) With a simple feature-matching loss at the target model's penultimate layer, heterogeneous VFMs become reliably stitchable across vision tasks. (3) For deep stitch points, the stitched model can surpass either constituent model at only a small inference overhead (for the stitch layer). Building on these findings, we further propose the VFM Stitch Tree (VST), which shares early layers across VFMs while retaining their later layers, yielding a controllable accuracy-latency trade-off for multimodal LLMs that often leverage multiple VFMs. Taken together, our study elevates stitching from a diagnostic probe to a practical recipe for integrating complementary VFM strengths and pinpointing where their representations align or diverge.

cs.CV↗
arXiv2026

X-Edit: Exact, Explicit, and Explainable Null-Space Editing for Medical Vision Transformers

Pre-trained Vision Transformers (ViTs) are increasingly deployed for medical image classification. However, correcting their inevitable failure cases in dynamic clinical scenarios poses a critical challenge. Conventional fine-tuning approaches inherently suffer from catastrophic forgetting, severely degrading previously acquired diagnostic capabilities. Such instability fundamentally compromises clinical safety. Addressing this vulnerability requires an active, controllable, and reliable intervention mechanism that is both theoretically grounded and inherently interpretable. To this end, we propose X-Edit (eXact, eXplicit, and eXplainable Editing), an efficient null-space model editing framework. X-Edit transitions the editing process from iterative gradient-based optimization to a theoretically grounded, closed-form solution. Specifically, we first explicitly localize the influential layers via causal tracing governing the erroneous prediction. Subsequently, we construct an orthogonal null-space projection matrix from a curated anchor set. By geometrically constraining the exact parameter update strictly within this null space, we provide mathematical guarantees that the intervention rectifies targeted errors without perturbing established diagnostic representations. Extensive evaluations on six medical imaging benchmarks demonstrate that X-Edit comprehensively suppresses catastrophic forgetting while achieving superior edit success rates. Our code is available at https://github.com/HenryLau7/X-Edit.

cs.CV↗
arXiv2026

Multi-Scale Generative Modeling with Heat Dissipation Flow Matching

Diffusion models are widely used in image generation, with most relying on noise-based corruption and denoising. A distinct branch instead uses blur as the main corruption, preserving better color budgets and multi-scale detail by providing multi-scale priors. However, blur-based models remain in SDE-based frameworks and are not integrated into ODE-based frameworks, such as Flow Matching (FM). Meanwhile, in the blur-based formulation, the classical inverse heat-dissipation (IHD) process faces an ill-posed challenge. Moreover, under the data-manifold assumption, regressing blurred images from high-dimensional noise (or velocity) space is also difficult. We propose Heat Dissipation Flow Matching (HDFM), which introduces a continuous blurred (heat-dissipation) process into FM to inject multi-scale priors. HDFM aligns an interpolated heat-dissipation path to address ill-posedness and adopts $x$-prediction to mitigate high-dimensional regression difficulty. Toy experiments and ablation studies show that HDFM consistently benefits from both blur and $x$-prediction. The performance of HDFM outperforms most baseline methods on all datasets.

cs.CV↗
arXiv2026

FastOCR: Dynamic Visual Fixation via KV Cache Pruning for Efficient Document Parsing

Vision-Language Models (VLMs) have shown strong promise on Optical Character Recognition (OCR), yet the sheer number of visual tokens required to encode dense documents incurs prohibitive inference cost. Existing pruning methods rely on physical eviction, e.g., permanently discarding visual tokens during the prefill stage. While effective for natural images, this strategy fundamentally breaks down on OCR, where virtually every visual token may correspond to a character or structural element, and any irreversible loss leads to catastrophic accuracy degradation. We observe that, although document images appear globally dense and seemingly unprunable, the model's attention to them is in fact temporally sparse: at each decoding step it concentrates on a small region that shifts gradually across steps, much as a human reader fixates on successive words rather than perceiving an entire page at once. Motivated by this Dynamic Visual Fixation phenomenon, we recast the intractable global pruning problem as a tractable local, dynamic one and propose FastOCR, a training-free framework with two complementary modules. Specifically, Focal-Guided Pruning identifies a small set of focal layers and selects the most task-relevant visual tokens from them at each step, while Cross-Step Fixation Reuse exploits the gradual shift of fixation to warm-start each step from the previous one. By dynamically adjusting which tokens are attended rather than evicting any from the cache, FastOCR avoids permanent information loss. Extensive experiments show that FastOCR serves as a plug-and-play acceleration module, generalizing consistently across five VLMs of varying sizes and architectures. On Qwen2.5-VL, FastOCR retains 98% of the unpruned model's accuracy while attending to only 5% of the visual tokens per decoding step, reducing attention latency by 3.0$\times$.

cs.CV↗
arXiv2026

OxyEcomBench: Benchmarking Multimodal Foundation Models across E-Commerce Ecosystems

LLMs and MLLMs have become indispensable tools across a wide range of applications. E-commerce, however, poses distinctive challenges -- including intricate domain knowledge, long-tail product evidence, heterogeneous visual data, and the interplay among multiple stakeholder roles -- that diverge substantially from the general world knowledge these models are primarily trained on, often causing a notable gap between their open-domain and e-commerce performance. To systematically quantify this gap, we introduce OxyEcomBench, a unified multimodal benchmark comprising approximately 6,300 high-quality instances for real-world bilingual Chinese--English e-commerce. Although several e-commerce benchmarks have been proposed, they typically adopt a single stakeholder perspective, target a narrow set of tasks, or address isolated challenges, making it difficult to holistically assess models' understanding of the full e-commerce pipeline. OxyEcomBench addresses these limitations by jointly covering platform operators, merchants, and customers across 6 capability aspects and 29 tasks, supporting text-only and mixed-modality inputs with single-image, multi-image, single-turn, and multi-turn configurations. All data is sourced from authentic e-commerce platforms and verified by domain experts. The benchmark further adopts a difficulty-aware design with a four-level P0--P3 rubric applied to all 29 tasks whose difficulty admits stable expert consensus, and rigorously prioritizes visually salient multimodal cases in which key evidence resides in images rather than text alone. Evaluations on 20 mainstream LLMs and MLLMs show that even the leading models attain modest performance and that performance gaps narrow on OxyEcomBench, suggesting that insufficient e-commerce-specific knowledge infusion mutes the advantages of advanced general-purpose models in this domain.

cs.DB↗
arXiv2026

IPAD-CLIP: Teaching CLIP to Detect Image Local Perceptual Artifacts

Current image quality assessment methods are heavily biased towards global distortions (e.g., noise, blur), neglecting local perceptual artifacts such as ghosting, lens flare, and moire effects. Although significant progress has been made in artifact removal, the fundamental problem of automatic artifact detection remains largely unexplored. In this paper, we formalize the Image Perceptual Artifact Detection (IPAD) task to address this gap. We contribute a benchmark dataset comprising 3,520 artifact images, including 520 real-captured and 3,000 synthetic samples, each paired with pixel-level masks across three representative artifact categories. The core challenge of IPAD lies in the localized, subtle, and semantically weak nature of these artifacts, which makes them prone to missed detection. To overcome this, we introduce IPAD-CLIP, a novel framework built upon CLIP that enhances artifact discrimination in both textual and visual spaces while preserving generalization capabilities. Our key insight is that local artifacts often exhibit strong correlations with specific semantic contexts. Accordingly, we learn artifact-aware text embeddings to explicitly model the object-artifact relationships, resulting in enhanced representations that clear differentiate between clean and artifact prompts. These text embeddings are then used as anchors to shift the visual encoder's attention from high-level semantics to subtle, low-level artifacts. Extensive experiments demonstrate that IPAD-CLIP offers a resource-efficient adaptation of CLIP for detection, significantly outperforming advanced image anomaly detection and manipulation detection methods on our benchmark. To the best of our knowledge, this is the first study addressing multi-class local perceptual artifact detection in terms of both dataset and model.

cs.CV↗
arXiv2026

Attention-Mamba: A Mamba-Enhanced Multi-Scale Parallel Inference Network for Medical Image Segmentation

U-shaped architectures have long dominated the field of medical image segmentation, while Transformers are widely employed for modeling long-range dependencies. The former typically handles scale variations implicitly by aggregating multi-level features, whereas the efficiency of the latter is constrained by its quadratic computational and memory complexity. In this work, we propose an effective alternative to traditional U-shaped architectures by constructing parallel branches at different levels to obtain multi-scale features and corresponding predictions. Furthermore, we enhance our network by integrating Mamba, a state space model that captures long-range dependencies with linear complexity. First, a dual-path architecture with lateral connections aggregates high-level semantic information and low-level spatial details at each branch. Then, we introduce a Recursive Alignment Module (RAM) that restores spatial details in low-resolution features through stepwise alignment, optimizing them for subsequent global feature learning and multi-scale fusion. We further build parallel Mamba branches upon aligned features to establish hierarchical global representations. Finally, we propose a Mamba-based attention mechanism for adaptive multi-scale prediction fusion; this mechanism utilizes Mamba to enhance information exchange across scales along both the channel and spatial dimensions. Experiments across three imaging modalities (MRI, CT, and dermoscopy) underscore the superior generalization of the proposed network. Compared to state-of-the-art 2D CNN, Transformer, and Mamba-based networks, our model achieves the highest segmentation performance on the Synapse, ACDC, ISIC-2018, and PH2 datasets while maintaining high efficiency, featuring the second-smallest parameters (14.05 M) and moderate computational complexity (8.94 GFLOPs).

cs.CV↗
arXiv2026

DialogueVPR: Towards Conversational Visual Place Recognition

Inspired by how humans communicate spatial information, language-guided geo-localization has gained significant traction for its intuitive and practical value. Despite this progress, most methods still rely on a static, one-shot retrieval paradigm, which fails to handle the ambiguity and incompleteness inherent in real-world natural language descriptions. We propose a paradigm shift to reasoning retrieval and introduce Dialogue Place Recognition (DlgPR), which casts localization as an interactive, dialogue-driven reasoning process. To support this new task, we present DlgQuest-Cities, the first large-scale dialogue-based benchmark for place recognition, and a unified reasoning framework that couples a cross-modal multi-level retriever with an intelligent questioner, DQ-pilot. DQ-pilot is trained in a curriculum: supervised fine-tuning on a curated DQ-cities-20k subset followed by reinforcement refinement on a harder DQ-cities-10k split via GRPO. Two task-aligned metrics guide learning: a Discriminative Difficulty Index (DDI) for curriculum sampling and a Positional Retrieval Gain (PRG) reward that directly measures retrieval improvement induced by a question. Experiments show this reasoning-based approach significantly outperforms baselines. The code and model are available at https://github.com/Graysonggg/DlgPR.

cs.AI↗
arXiv2026

ZScribbleSeg: A comprehensive segmentation framework with modeling of efficient annotation and maximization of scribble supervision

Curating fully annotated datasets for medical image segmentation is labour-intensive and expertise-demanding. To alleviate this problem, prior studies have explored scribble annotations for weakly supervised segmentation. Existing solutions mainly compute losses on annotated areas and generate pseudo labels by propagating annotations to adjacent regions. However, these methods often suffer from inaccurate and unrealistic segmentations due to insufficient supervision and incomplete shape information. In contrast, we first investigate the principle of good scribble annotations, which leads to efficient scribble forms via supervision maximization and randomness simulation. We further introduce regularization terms to encode the spatial relationship and the shape constraints, where the EM algorithm is utilized to estimate the mixture ratios of label classes. These ratios are critical in identifying the unlabeled pixels for each class and correcting erroneous predictions, thus the accurate estimation lays the foundation for the incorporation of spatial prior. Finally, we integrate the efficient scribble supervision with the prior into a framework, referred to as ZScribbleSeg, and apply it to multiple scenarios. Leveraging only scribble annotations, ZScribbleSeg achieves competitive performance on six segmentation tasks including ACDC, MSCMRseg, BTCV, MyoPS, Decathlon-BrainTumor and Decathlon-Prostate. Our code will be released via https://github.com/DLwbm123/ZScribbleSeg.

cs.CV↗