Searcharxiv⌕ Search

arXiv subjects

Yue Yang

Publications and source records attributed to Yue Yang.

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

Anisotropic Modality Align

Training multimodal large language models has long been limited by the scarcity of high-quality paired multimodal data. Recent studies show that the shared representation space of pretrained multimodal contrastive models can serve as a bridge, enabling models to perform multimodal training with unimodal data. However, the key premise of this paradigm remains insufficiently understood: can representations from different modalities be reliably interchanged? The core obstacle lies in the persistent Modality Gap in the shared space. In this work, we revisit the geometric nature of the modality gap. We find that modality representations already share compatible dominant semantic geometry. What truly hinders modality interchangeability is not a simple global shift, but an anisotropic residual structure concentrated along a small number of dominant directions. Based on this finding, we further propose the principle of anisotropic modality gap alignment: effective modality alignment should align with the target-modality distribution while preserving the semantic structure of the source modality. Guided by this principle, we propose an anisotropic geometric correction framework, AnisoAlign, for unpaired modality alignment. This framework leverages the internal geometric prior of the target modality and performs bounded correction on source-modality representations, thereby constructing substitute representations in the target modality. Experiments confirm its benefits in both geometric diagnostics and text-only MLLM training. Overall, this work recasts the modality gap from an empirical observation into a correctable, structured geometric phenomenon and provides a new representation alignment perspective for training multimodal models with unimodal data.

cs.MM↗
arXiv2026

Geometric encoding of turbulence for end-to-end quantum simulation

Multiscale organization is a hallmark of fluid turbulence in aerospace, energy, and transport systems. While quantum computing promises exponential speedups for solving the evolution equations governing flow fields, this potential is fundamentally hindered by the quantum state preparation bottleneck, the prohibitive cost of loading classical complex data into quantum states. Here, we overcome this barrier by introducing a physics-informed, three-stage geometric encoding method "turbuloscope", which efficiently generates turbulent fields relevant to high-Reynolds-number engineering flows. Rather than brute-force data loading, our approach acts as a kaleidoscope, leveraging the multiscale structures of turbulence. We capture scale-invariant self-similarity via a hyperplane approximation in high-dimensional feature space, and utilize the Hopf fibration to map quantum observables directly onto vortex tubes, the fundamental building blocks of turbulence that control mixing, drag, and heat transfer in mechanical systems. Remarkably, the algorithm requires no ancillary qubits, utilizes a linear-depth quantum circuit, and scales logarithmically with the Reynolds number, an exponential speedup compared to classical methods. We demonstrate the power of this method by generating an instantaneous turbulent field at a high Reynolds number of 35,000 across over one billion grid points using only 30 qubits, reproducing Kolmogorov's 5/3 energy spectrum, tangled vortex structures, and strong intermittency. This asymptotically optimal approach not only signals a near-term pathway to practical quantum advantage in engineering simulation, but establishes a scalable foundation for the quantum simulation of broad multiscale systems.

quant-ph↗
arXiv2026

Musical Score Understanding Benchmark: Evaluating Large Language Models' Comprehension of Complete Musical Scores

Understanding complete musical scores entails integrated reasoning over pitch, rhythm, harmony, and large-scale structure, yet the ability of Large Language Models and Vision--Language Models to interpret full musical notation remains insufficiently examined. We introduce Musical Score Understanding Benchmark (MSU-Bench), a human-curated benchmark for score-level musical understanding across textual (ABC notation) and visual (PDF) modalities. MSU-Bench contains 1,800 generative question-answer pairs from works by Bach, Beethoven, Chopin, Debussy, and others, organised into four levels of increasing difficulty, ranging from onset information to texture and form. Evaluations of more than fifteen state-of-the-art models, in both zero-shot and fine-tuned settings, reveal pronounced modality gaps, unstable level-wise performance, and challenges in maintaining multilevel correctness. Fine-tuning substantially improves results across modalities while preserving general knowledge, positioning MSU-Bench as a robust foundation for future research in multimodal reasoning. The benchmark and code are available at https://github.com/Congren-Dai/MSU-Bench.

cs.SD↗
arXiv2026

Seeing Through Experts Eyes A Foundational Vision Language Model Trained on Radiologists Gaze and Reasoning

Large scale vision language models have shown promise in automating chest Xray interpretation, yet their clinical utility remains limited by a gap between model outputs and radiologist reasoning. Most systems optimize for semantic information without emulating how experts visually examine medical images, often overlooking critical findings or diverging from established diagnostic workflows. Radiologists follow structured protocols (e.g., the ABCDEF approach) that ensure all clinically relevant regions are systematically examined, reducing missed findings and supporting reliable diagnostic reasoning. We introduce GazeX, a vision language model that leverages radiologists' eye tracking data as a behavioral prior to model expert diagnostic reasoning. By incorporating gaze trajectories and fixation patterns into pretraining, GazeX learns to follow the spatial and temporal structure of radiologist attention and integrates observations in a clinically meaningful sequence. Using a curated dataset of over 30,000 gaze key frames from five radiologists, we demonstrate that GazeX produces more accurate, interpretable, and expert consistent outputs across radiology report generation, disease grounding, and visual question answering, utilizing 231,835 radiographic studies, 780,014 question answer pairs, and 1,162 image sentence pairs with bounding boxes. Unlike autonomous reporting systems, GazeX produces verifiable evidence artifacts, including inspection trajectories and finding linked localized regions, enabling efficient human verification and safe human AI collaboration. Learning through expert eyes provides a practical route toward more trustworthy, explainable, and diagnostically robust AI systems for radiology and beyond.

cs.AI↗
arXiv2026

MolmoWeb: Open Visual Web Agent and Open Data for the Open Web

Web agents--autonomous systems that navigate and execute tasks on the web on behalf of users--have the potential to transform how people interact with the digital world. However, the most capable web agents today rely on proprietary models with undisclosed training data and recipes, limiting scientific understanding, reproducibility, and community-driven progress. We believe agents for the open web should be built in the open. To this end, we introduce (1) MolmoWebMix, a large and diverse mixture of browser task demonstrations and web-GUI perception data and (2) MolmoWeb, a family of fully open multimodal web agents. Specifically, MolmoWebMix combines over 100K synthetic task trajectories from multiple complementary generation pipelines with 30K+ human demonstrations, atomic web-skill trajectories, and GUI perception data, including referring expression grounding and screenshot question answering. MolmoWeb agents operate as instruction-conditioned visual-language action policies: given a task instruction and a webpage screenshot, they predict the next browser action, requiring no access to HTML, accessibility trees, or specialized APIs. Available in 4B and 8B size, on browser-use benchmarks like WebVoyager, Online-Mind2Web, and DeepShop, MolmoWeb agents achieve state-of-the-art results outperforming similar scale open-weight-only models such as Fara-7B, UI-Tars-1.5-7B, and Holo1-7B. MolmoWeb-8B also surpasses set-of-marks (SoM) agents built on much larger closed frontier models like GPT-4o. We further demonstrate consistent gains through test-time scaling via parallel rollouts with best-of-N selection, achieving 94.7% and 60.5% pass@4 (compared to 78.2% and 35.3% pass@1) on WebVoyager and Online-Mind2Web respectively. We will release model checkpoints, training data, code, and a unified evaluation harness to enable reproducibility and accelerate open research on web agents.

cs.CV↗
arXiv2026

Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Grounding

Today's strongest video-language models (VLMs) remain proprietary. The strongest open-weight models either rely on synthetic data from proprietary VLMs, effectively distilling from them, or do not disclose their training data or recipe. As a result, the open-source community lacks the foundations needed to improve on the state-of-the-art video (and image) language models. Crucially, many downstream applications require more than just high-level video understanding; they require grounding -- either by pointing or by tracking in pixels. Even proprietary models lack this capability. We present Molmo2, a new family of VLMs that are state-of-the-art among open-source models and demonstrate exceptional new capabilities in point-driven grounding in single image, multi-image, and video tasks. Our key contribution is a collection of 7 new video datasets and 2 multi-image datasets, including a dataset of highly detailed video captions for pre-training, a free-form video Q&A dataset for fine-tuning, a new object tracking dataset with complex queries, and an innovative new video pointing dataset, all collected without the use of closed VLMs. We also present a training recipe for this data utilizing an efficient packing and message-tree encoding scheme, and show bi-directional attention on vision tokens and a novel token-weight strategy improves performance. Our best-in-class 8B model outperforms others in the class of open weight and data models on short videos, counting, and captioning, and is competitive on long-videos. On video-grounding Molmo2 significantly outperforms existing open-weight models like Qwen3-VL (35.5 vs 29.6 accuracy on video counting) and surpasses proprietary models like Gemini 3 Pro on some tasks (38.4 vs 20.0 F1 on video pointing and 56.2 vs 41.1 J&F on video tracking).

cs.CV↗
arXiv2026

MolmoPoint: Better Pointing for VLMs with Grounding Tokens

Grounding has become a fundamental capability of vision-language models (VLMs). Most existing VLMs point by generating coordinates as part of their text output, which requires learning a complicated coordinate system and results in a high token count. Instead, we propose a more intuitive pointing mechanism that directly selects the visual tokens that contain the target concept. Our model generates a special pointing token that cross-attends to the input image or video tokens and selects the appropriate one. To make this model more fine-grained, we follow these pointing tokens with an additional special token that selects a fine-grained subpatch within the initially selected region, and then a third token that specifies a location within that subpatch. We further show that performance improves by generating points sequentially in a consistent order, encoding the relative position of the previously selected point, and including a special no-more-points class when selecting visual tokens. Using this method, we set a new state-of-the-art on image pointing (70.7% on PointBench), set a new state-of-the-art among fully open models on GUI pointing (61.1% on ScreenSpotPro), and improve video pointing (59.1% human preference win rate vs. a text coordinate baseline) and tracking (+6.3% gain on Molmo2Track). We additionally show that our method achieves much higher sample efficiency and discuss the qualitative differences that emerge from this design change.

cs.CV↗
arXiv2026

From Coarse to Continuous: Progressive Refinement Implicit Neural Representation for Motion-Robust Anisotropic MRI Reconstruction

In motion-robust magnetic resonance imaging (MRI), slice-to-volume reconstruction is critical for recovering anatomically consistent 3D brain volumes from 2D slices, especially under accelerated acquisitions or patient motion. However, this task remains challenging due to hierarchical structural disruptions. It includes local detail loss from k-space undersampling, global structural aliasing caused by motion, and volumetric anisotropy. Therefore, we propose a progressive refinement implicit neural representation (PR-INR) framework. Our PR-INR unifies motion correction, structural refinement, and volumetric synthesis within a geometry-aware coordinate space. Specifically, a motion-aware diffusion module is first employed to generate coarse volumetric reconstructions that suppress motion artifacts and preserve global anatomical structures. Then, we introduce an implicit detail restoration module that performs residual refinement by aligning spatial coordinates with visual features. It corrects local structures and enhances boundary precision. Further, a voxel continuous-aware representation module represents the image as a continuous function over 3D coordinates. It enables accurate inter-slice completion and high-frequency detail recovery. We evaluate PR-INR on five public MRI datasets under various motion conditions (3% and 5% displacement), undersampling rates (4x and 8x) and slice resolutions (scale = 5). Experimental results demonstrate that PR-INR outperforms state-of-the-art methods in both quantitative reconstruction metrics and visual quality. It further shows generalization and robustness across diverse unseen domains.

eess.IV↗
arXiv2026

Unified Spatio-Temporal Token Scoring for Efficient Video VLMs

Token pruning is essential for enhancing the computational efficiency of vision-language models (VLMs), particularly for video-based tasks where temporal redundancy is prevalent. Prior approaches typically prune tokens either (1) within the vision transformer (ViT) exclusively for unimodal perception tasks such as action recognition and object segmentation, without adapting to downstream vision-language tasks; or (2) only within the LLM while leaving the ViT output intact, often requiring complex text-conditioned token selection mechanisms. In this paper, we introduce Spatio-Temporal Token Scoring (STTS), a simple and lightweight module that prunes vision tokens across both the ViT and the LLM without text conditioning or token merging, and is fully compatible with end-to-end training. By learning how to score temporally via an auxiliary loss and spatially via LLM downstream gradients, aided by our efficient packing algorithm, STTS prunes 50% of vision tokens throughout the entire architecture, resulting in a 62% improvement in efficiency during both training and inference with only a 0.7% drop in average performance across 13 short and long video QA tasks. Efficiency gains increase with more sampled frames per video. Applying test-time scaling for long-video QA further yields performance gains of 0.5-1% compared to the baseline. Overall, STTS represents a novel, simple yet effective technique for unified, architecture-wide vision token pruning.

cs.CV↗
arXiv2026

EMMA: Generalizing Real-World Robot Manipulation via Generative Visual Transfer

The generalization of vision-language-action (VLA) models heavily relies on diverse training data. However, acquiring large-scale data for robot manipulation across varied object appearances is costly and labor-intensive. To address this limitation, we introduce Embodied Manipulation Media Adaptation (EMMA), a framework for augmenting VLA policies that combines a generative data engine with an effective training pipeline. We introduce DreamTransfer, a diffusion Transformer-based architecture for generating multi-view consistent and geometrically grounded embodied manipulation videos. DreamTransfer enables visual editing of robot videos through prompts, allowing for changes to the foreground, background, and lighting while preserving their 3D structure and geometric validity. We also utilize a hybrid training set of real and generated data and propose AdaMix to enhance the training process. AdaMix is a training strategy that adaptively weights samples according to policy performance to emphasize challenging samples. Comprehensive evaluations demonstrate that videos created by DreamTransfer yield substantial improvements over previous video generation techniques in multi-view consistency, geometric accuracy, and text-conditioning precision. We conduct extensive evaluations with a total of more than 1800 trials in both simulated and real-world robotic environments. In real-world robotic tasks with zero-shot visual settings, our framework achieves a relative performance increase of over 92% compared to training with real data alone, and improves by an additional 17% with AdaMix, demonstrating its efficacy in enhancing policy generalization.

cs.AI↗
arXiv2026

Quantum computing of nonlinear reacting flows via the probability density function method

Quantum computing offers the promise of speedups for scientific computations, but its application to reacting flows is hindered by nonlinear source terms, the challenges of time-dependent simulations, and the difficulty of extracting meaningful physical quantities from quantum states. We employ a probability density function (PDF) formulation to transform the nonlinear reacting-flow governing equations into high-dimensional linear ones. The entire temporal evolution is then solved as a single large linear system using the history state method, which avoids the measurement bottleneck of conventional time-marching schemes and fully leverages the advantages of quantum linear system algorithms. To extract the quantity of interest from the resulting quantum state, we develop an efficient algorithm to measure the statistical moments of the PDF, bypassing the need for costly full-state tomography. A computational complexity analysis shows that the measurement algorithm achieves a complexity polynomial in the logarithm of the system size using low-order polynomial approximations, compared to the exponential cost of the exact operator, thereby retaining the quantum advantage gained from solving the linear system. We validate the framework in two stages: an a priori test confirms the accuracy of the measurement algorithm on beta distributions with known analytical moments, and a perfectly stirred reactor simulation demonstrates the capability to capture the PDF evolution and statistics of a nonlinear reactive system. This work establishes a pathway for applying quantum computing to nonlinear reacting flows.

quant-ph↗
arXiv2026

Airfoil shape optimization via coherent Ising machine

Airfoil shape optimization presents a challenge where classical solvers frequently struggle with computational efficiency and local minima. In the promising paradigm of quantum computing, the coherent Ising machine (CIM), a specialized physical solver, offers acceleration capabilities. However, its native discrete binary architecture restricts the application in aerodynamic design. To bridge this gap, we propose a comprehensive framework that translates airfoil shape optimization into hardware-compliant quadratic unconstrained binary optimization formulations. We integrate high-order response surface models via the Rosenberg order reduction, enabling the CIM to capture strong nonlinearities in the aerodynamic performance response. Furthermore, we introduce a block-diagonal scalarization strategy that compose trade-off scenarios into a single optimization. Validated on the NACA 4-digit airfoil series using CIM hardware with 615 spins, the framework successfully locates the global optimum with a computational speedup of three orders of magnitude compared to the classical simulated annealing. The parallel embedding capacity allows for the extraction of an entire optimal Pareto front in a single hardware execution. This work demonstrates a viable, quantum-enhanced paradigm for engineering optimization.

quant-ph↗
arXiv2026

3D Wavelet-Based Structural Priors for Controlled Diffusion in Whole-Body Low-Dose PET Denoising

Low-dose Positron Emission Tomography (PET) imaging reduces patient radiation exposure but suffers from increased noise that degrades image quality and diagnostic reliability. Although diffusion models have demonstrated strong denoising capability, their stochastic nature makes it challenging to enforce anatomically consistent structures, particularly in low signal-to-noise regimes and volumetric whole-body imaging. We propose Wavelet-Conditioned ControlNet (WCC-Net), a fully 3D diffusion-based framework that introduces explicit frequency-domain structural priors via wavelet representations to guide volumetric PET denoising. By injecting wavelet-based structural guidance into a frozen pretrained diffusion backbone through a lightweight control branch, WCC-Net decouples anatomical structure from noise while preserving generative expressiveness and 3D structural continuity. Extensive experiments demonstrate that WCC-Net consistently outperforms CNN-, GAN-, and diffusion-based baselines. On the internal 1/20-dose test set, WCC-Net improves PSNR by +1.21 dB and SSIM by +0.008 over a strong diffusion baseline, while reducing structural distortion (GMSD) and intensity error (NMAE). Moreover, WCC-Net generalizes robustly to unseen dose levels (1/50 and 1/4), achieving superior quantitative performance and improved volumetric anatomical consistency.

cs.CV↗
arXiv2026

LiLo-VLA: Compositional Long-Horizon Manipulation via Linked Object-Centric Policies

General-purpose robots must master long-horizon manipulation, defined as tasks involving multiple kinematic structure changes (e.g., attaching or detaching objects) in unstructured environments. While Vision-Language-Action (VLA) models offer the potential to master diverse atomic skills, they struggle with the combinatorial complexity of sequencing them and are prone to cascading failures due to environmental sensitivity. To address these challenges, we propose LiLo-VLA (Linked Local VLA), a modular framework capable of zero-shot generalization to novel long-horizon tasks without ever being trained on them. Our approach decouples transport from interaction: a Reaching Module handles global motion, while an Interaction Module employs an object-centric VLA to process isolated objects of interest, ensuring robustness against irrelevant visual features and invariance to spatial configurations. Crucially, this modularity facilitates robust failure recovery through dynamic replanning and skill reuse, effectively mitigating the cascading errors common in end-to-end approaches. We introduce a 21-task simulation benchmark consisting of two challenging suites: LIBERO-Long++ and Ultra-Long. In these simulations, LiLo-VLA achieves a 69% average success rate, outperforming Pi0.5 by 41% and OpenVLA-OFT by 67%. Furthermore, real-world evaluations across 8 long-horizon tasks demonstrate an average success rate of 85%. Project page: https://yy-gx.github.io/LiLo-VLA/.

cs.RO↗
arXiv2026

Connecting the Dots: Surfacing Structure in Documents through AI-Generated Cross-Modal Links

Understanding information-dense documents like recipes and scientific papers requires readers to find, interpret, and connect details scattered across text, figures, tables, and other visual elements. These documents are often long and filled with specialized terminology, hindering the ability to locate relevant information or piece together related ideas. Existing tools offer limited support for synthesizing information across media types. As a result, understanding complex material remains cognitively demanding. This paper presents a framework for fine-grained integration of information in complex documents. We instantiate the framework in an augmented reading interface, which populates a scientific paper with clickable points on figures, interactive highlights in the body text, and a persistent reference panel for accessing consolidated details without manual scrolling. In a controlled between-subjects study, we find that participants who read the paper with our tool achieved significantly higher scores on a reading quiz without evidence of increased time to completion or cognitive load. Fine-grained integration provides a systematic way of revealing relationships within a document, supporting engagement with complex, information-dense materials.

cs.HC↗
arXiv2026

REDSearcher: A Scalable and Cost-Efficient Framework for Long-Horizon Search Agents

Large language models are transitioning from generalpurpose knowledge engines to realworld problem solvers, yet optimizing them for deep search tasks remains challenging. The central bottleneck lies in the extreme sparsity of highquality search trajectories and reward signals, arising from the difficulty of scalable longhorizon task construction and the high cost of interactionheavy rollouts involving external tool calls. To address these challenges, we propose REDSearcher, a unified framework that codesigns complex task synthesis, midtraining, and posttraining for scalable searchagent optimization. Specifically, REDSearcher introduces the following improvements: (1) We frame task synthesis as a dualconstrained optimization, where task difficulty is precisely governed by graph topology and evidence dispersion, allowing scalable generation of complex, highquality tasks. (2) We introduce toolaugmented queries to encourage proactive tool use rather than passive recall.(3) During midtraining, we strengthen core atomic capabilities knowledge, planning, and function calling substantially reducing the cost of collecting highquality trajectories for downstream training. (4) We build a local simulated environment that enables rapid, lowcost algorithmic iteration for reinforcement learning experiments. Across both textonly and multimodal searchagent benchmarks, our approach achieves stateoftheart performance. To facilitate future research on longhorizon search agents, we will release 10K highquality complex text search trajectories, 5K multimodal trajectories and 1K text RL query set, and together with code and model checkpoints.

cs.AI↗
arXiv2026

Detecting RLVR Training Data via Structural Convergence of Reasoning

Reinforcement learning with verifiable rewards (RLVR) is central to training modern reasoning models, but the undisclosed training data raises concerns about benchmark contamination. Unlike pretraining methods, which optimize models using token-level probabilities, RLVR fine-tunes models based on reward feedback from self-generated reasoning trajectories, making conventional likelihood-based detection methods less effective. We show that RLVR induces a distinctive behavioral signature: prompts encountered during RLVR training result in more rigid and similar generations, while unseen prompts retain greater diversity. We introduce Min-$k$NN Distance, a simple black-box detector that quantifies this collapse by sampling multiple completions for a given prompt and computing the average of the $k$ smallest nearest-neighbor edit distances. Min-$k$NN Distance requires no access to the reference model or token probabilities. Experiments across multiple RLVR-trained reasoning models show that Min-$k$NN Distance reliably distinguishes RL-seen examples from unseen ones and outperforms existing membership inference and RL contamination detection baselines.

cs.AI↗
arXiv2026

Optimizing Automated Picking Systems in Warehouse Robots Using Machine Learning

With the rapid growth of global e-commerce, the demand for automation in the logistics industry is increasing. This study focuses on automated picking systems in warehouses, utilizing deep learning and reinforcement learning technologies to enhance picking efficiency and accuracy while reducing system failure rates. Through empirical analysis, we demonstrate the effectiveness of these technologies in improving robot picking performance and adaptability to complex environments. The results show that the integrated machine learning model significantly outperforms traditional methods, effectively addressing the challenges of peak order processing, reducing operational errors, and improving overall logistics efficiency. Additionally, by analyzing environmental factors, this study further optimizes system design to ensure efficient and stable operation under variable conditions. This research not only provides innovative solutions for logistics automation but also offers a theoretical and empirical foundation for future technological development and application.

cs.RO↗