Searcharxiv⌕ Search

arXiv subjects

Yu Liu

Publications and source records attributed to Yu Liu.

At least 307 records · Page 17Linked to original sources
arXiv2026

GDRO: Group-level Reward Post-training Suitable for Diffusion Models

Recent advancements adopt online reinforcement learning (RL) from LLMs to text-to-image rectified flow diffusion models for reward alignment. The use of group-level rewards successfully aligns the model with the targeted reward. However, it faces challenges including low efficiency, dependency on stochastic samplers, and reward hacking. The problem is that rectified flow models are fundamentally different from LLMs: 1) For efficiency, online image sampling takes much more time and dominates the time of training. 2) For stochasticity, rectified flow is deterministic once the initial noise is fixed. Aiming at these problems and inspired by the effects of group-level rewards from LLMs, we design Group-level Direct Reward Optimization (GDRO). GDRO is a new post-training paradigm for group-level reward alignment that combines the characteristics of rectified flow models. Through rigorous theoretical analysis, we point out that GDRO supports full offline training that saves the large time cost for image rollout sampling. Also, it is diffusion-sampler-independent, which eliminates the need for the ODE-to-SDE approximation to obtain stochasticity. We also empirically study the reward hacking trap that may mislead the evaluation, and involve this factor in the evaluation using a corrected score that not only considers the original evaluation reward but also the trend of reward hacking. Extensive experiments demonstrate that GDRO effectively and efficiently improves the reward score of the diffusion model through group-wise offline optimization across the OCR and GenEval tasks, while demonstrating strong stability and robustness in mitigating reward hacking.

cs.LG↗
arXiv2025

FDP: A Frequency-Decomposition Preprocessing Pipeline for Unsupervised Anomaly Detection in Brain MRI

Due to the diversity of brain anatomy and the scarcity of annotated data, supervised anomaly detection for brain MRI remains challenging, driving the development of unsupervised anomaly detection (UAD) approaches. Current UAD methods typically utilize artificially generated noise perturbations on healthy MRIs to train generative models for normal anatomy reconstruction, enabling anomaly detection via residual maps. However, such simulated anomalies lack the biophysical fidelity and morphological complexity characteristic of true clinical lesions. To advance UAD in brain MRI, we conduct the first systematic frequency-domain analysis of pathological signatures, revealing two key properties: (1) anomalies exhibit unique frequency patterns distinguishable from normal anatomy, and (2) low-frequency signals maintain consistent representations across healthy scans. These insights motivate our Frequency-Decomposition Preprocessing (FDP) framework, the first UAD method to leverage frequency-domain reconstruction for simultaneous pathology suppression and anatomical preservation. FDP can integrate seamlessly with existing anomaly simulation techniques, consistently enhancing detection performance across diverse architectures while maintaining diagnostic fidelity. Experimental results demonstrate that FDP consistently improves anomaly detection performance when integrated with existing methods. Notably, FDP achieves a 17.63% increase in DICE score with LDM while maintaining robust improvements across multiple baselines. The code is available at https://github.com/ls1rius/MRI_FDP.

cs.CV↗
arXiv2025

MME-CC: A Challenging Multi-Modal Evaluation Benchmark of Cognitive Capacity

As reasoning models scale rapidly, the essential role of multimodality in human cognition has come into sharp relief, driving a growing need to probe vision-centric cognitive behaviors. Yet, existing multimodal benchmarks either overemphasize textual reasoning or fall short of systematically capturing vision-centric cognitive behaviors, leaving the cognitive capacity of MLLMs insufficiently assessed. To address this limitation, we introduce MME-CC (Multi-Modal Evaluation benchmark of Cognitive Capacity), a vision-grounded benchmark that organizes 11 representative reasoning tasks into three fundamental categories of visual information: spatial, geometric, and knowledge-based reasoning, and provides fine-grained analyses of MLLMs' cognitive capacity across these dimensions. Based on MME-CC, we conduct extensive experiments over 16 representative MLLMs. Our study reveals that closed-source models currently lead overall (e.g., 42.66 for Gemini-2.5-Pro vs. 30.45 for GLM-4.5V), while spatial and geometric reasoning remain broadly weak (less than or equal to 30%). We further identify common error patterns, including orientation mistakes, fragile cross-view identity persistence, and poor adherence to counterfactual instructions, and observe that Chain-of-Thought typically follows a three-stage process (extract -> reason -> verify) with heavy reliance on visual extraction. We hope this work catalyzes a shift toward treating the cognitive capacity of MLLMs as central to both evaluation and model design.

cs.CL↗
arXiv2025

Spatial-aware Symmetric Alignment for Text-guided Medical Image Segmentation

Text-guided Medical Image Segmentation has shown considerable promise for medical image segmentation, with rich clinical text serving as an effective supplement for scarce data. However, current methods have two key bottlenecks. On one hand, they struggle to process diagnostic and descriptive texts simultaneously, making it difficult to identify lesions and establish associations with image regions. On the other hand, existing approaches focus on lesions description and fail to capture positional constraints, leading to critical deviations. Specifically, with the text "in the left lower lung", the segmentation results may incorrectly cover both sides of the lung. To address the limitations, we propose the Spatial-aware Symmetric Alignment (SSA) framework to enhance the capacity of referring hybrid medical texts consisting of locational, descriptive, and diagnostic information. Specifically, we propose symmetric optimal transport alignment mechanism to strengthen the associations between image regions and multiple relevant expressions, which establishes bi-directional fine-grained multimodal correspondences. In addition, we devise a composite directional guidance strategy that explicitly introduces spatial constraints in the text by constructing region-level guidance masks. Extensive experiments on public benchmarks demonstrate that SSA achieves state-of-the-art (SOTA) performance, particularly in accurately segmenting lesions characterized by spatial relational constraints.

cs.CV↗
arXiv2025

Rewards-based image analysis in microscopy

Imaging and hyperspectral data analysis is central to progress across biology, medicine, chemistry, and physics. The core challenge lies in converting high-resolution or high-dimensional datasets into interpretable representations that enable insight into the underlying physical or chemical properties of a system. Traditional analysis relies on expert-designed, multistep workflows, such as denoising, feature extraction, clustering, dimensionality reduction, and physics-based deconvolution, or on machine learning (ML) methods that accelerate individual steps. Both approaches, however, typically demand significant human intervention, including hyperparameter tuning and data labeling. Achieving the next level of autonomy in scientific imaging requires designing effective reward-based workflows that guide algorithms toward best data representation for human or automated decision-making. Here, we discuss recent advances in reward-based workflows for image analysis, which capture key elements of human reasoning and exhibit strong transferability across various tasks. We highlight how reward-driven approaches enable a shift from supervised black-box models toward explainable, unsupervised optimization on the examples of Scanning Probe and Electron Microscopies. Such reward-based frameworks are promising for a broad range of applications, including classification, regression, structure-property mapping, and general hyperspectral data processing.

eess.IV↗
arXiv2025

Overcoming frequency resolution limits using a solid-state spin quantum sensor

The ability to determine precisely the separation of two frequencies is fundamental to spectroscopy, yet the resolution limit poses a critical challenge: distinguishing two incoherent signals becomes impossible when their frequencies are sufficiently close. Here, we demonstrate a simple and powerful approach, dubbed {\it superresolution quantum sensing}, which experimentally resolves two nearly identical incoherent signals using a solid-state spin quantum sensor. By carefully choosing interrogation times that satisfy the superresolution condition, we eliminate quantum projection noise, overcoming the vanishing distinguishability of signals with near-identical frequencies. This leads to improved resolution, which scales as $t^{-2}$ in comparison to the standard $t^{-1}$ scaling. Together with a greatly reduced classical readout noise assisted by a nuclear spin, we are able to achieve sub-kHz resolution with a signal detection time of 80 microseconds. Our results highlight the potential of quantum sensing to overcome conventional frequency resolution limitations, with broad implications for precision measurements.

quant-ph↗
arXiv2025

IndoorUAV: Benchmarking Vision-Language UAV Navigation in Continuous Indoor Environments

Vision-Language Navigation (VLN) enables agents to navigate in complex environments by following natural language instructions grounded in visual observations. Although most existing work has focused on ground-based robots or outdoor Unmanned Aerial Vehicles (UAVs), indoor UAV-based VLN remains underexplored, despite its relevance to real-world applications such as inspection, delivery, and search-and-rescue in confined spaces. To bridge this gap, we introduce \textbf{IndoorUAV}, a novel benchmark and method specifically tailored for VLN with indoor UAVs. We begin by curating over 1,000 diverse and structurally rich 3D indoor scenes from the Habitat simulator. Within these environments, we simulate realistic UAV flight dynamics to collect diverse 3D navigation trajectories manually, further enriched through data augmentation techniques. Furthermore, we design an automated annotation pipeline to generate natural language instructions of varying granularity for each trajectory. This process yields over 16,000 high-quality trajectories, comprising the \textbf{IndoorUAV-VLN} subset, which focuses on long-horizon VLN. To support short-horizon planning, we segment long trajectories into sub-trajectories by selecting semantically salient keyframes and regenerating concise instructions, forming the \textbf{IndoorUAV-VLA} subset. Finally, we introduce \textbf{IndoorUAV-Agent}, a novel navigation model designed for our benchmark, leveraging task decomposition and multimodal reasoning. We hope IndoorUAV serves as a valuable resource to advance research on vision-language embodied AI in the indoor aerial navigation domain.

cs.RO↗
arXiv2025

Resonant ALP-Portal Dark Matter Annihilation as a Solution to the $B^{\pm} \to K^{\pm} ν\barν$ Excess

The Belle II collaboration recently reported a $2.7σ$ excess in the rare decay $B^\pm \to K^\pm ν\barν$, potentially signaling new physics. We propose an axion-like particle (ALP)-portal dark matter (DM) framework to explain this anomaly while satisfying the observed DM relic abundance. By invoking a resonant annihilation mechanism ($m_a \sim 2m_χ$), we demonstrate that the ALP-mediated interactions between the Standard Model and DM sectors simultaneously account for the $B^\pm \to K^\pm ν\barν$ anomaly and thermal freeze-out dynamics. Two distinct scenarios-long-lived ALPs decaying outside detectors (displaced diphotons) and ALPs decaying invisibly to DM pairs (missing energy)-are examined. While the displaced diphotons scenario is excluded by kaon decay bounds ($K^\pm \to π^\pm + \text{inv.}$), the invisible decay channel remains unconstrained and aligns with Belle II's missing energy signature. Using the coupled Boltzmann equation formalism, we rigorously incorporate early kinetic decoupling effects, revealing deviations up to a factor of 20 from traditional relic density predictions in resonance regions. For the missing energy scenario, the viable parameter space features ALP-SM and ALP-DM couplings: $g_{aWW}(g_{aγγ}) \in (7.13 \times 10^{-5} - 9.60 \times 10^{-5})\, \text{GeV}^{-1}$ (from $B^\pm \to K^\pm a$) and $g_{aχχ} \in (7.12\times10^{-5} - 7.73\times 10^{-3})\, \text{GeV}^{-1}$ (for resonant annihilation), accommodating ALP masses $m_a \in (0.6, 4.8)\, \text{GeV}$. Therefore, this work establishes the ALP portal as a viable bridge between the $B^\pm \to K^\pm ν\barν$ anomaly and thermal DM production, emphasizing precision calculations of thermal decoupling in resonance regimes.

hep-ph↗
arXiv2025

SWE-Bench++: A Framework for the Scalable Generation of Software Engineering Benchmarks from Open-Source Repositories

Benchmarks like SWE-bench have standardized the evaluation of Large Language Models (LLMs) on repository-level software engineering tasks. However, these efforts remain limited by manual curation, static datasets, and a focus on Python-based bug fixes. We introduce SWE-Bench++, an automated framework that generates repository-level coding tasks from open-source GitHub projects. Unlike synthetic approaches, our pipeline harvests live pull requests to cover both bug fixes and feature requests across 11 languages. SWE-Bench++ turns GitHub pull requests (PRs) into reproducible, execution-based tasks via four stages: programmatic sourcing, environment synthesis, test oracle extraction, and quality assurance. A final hint-guided trajectory synthesis step converts instances that strong models fail on into training trajectories. Our initial benchmark consists of 11,133 instances from 3,971 repositories across 11 languages. On a subset of 1,782 instances of this benchmark, today's strongest models perform as follows: claude-sonnet-4.5 achieves 36.20% pass@10, gpt-5-2025-08-07 34.57%, gemini/gemini-2.5-pro 24.92%, and gpt-4o 16.89%. We further demonstrate the utility of our dataset by showing that fine-tuning on SWE-Bench++ instances yields measurable improvements on the SWE-bench Multilingual benchmark. SWE-Bench++ provides a scalable, multilingual benchmark for evaluating and improving repository-level code generation.

cs.SE↗
arXiv2025

Interplay of Defects and the Charge Density Wave State in Hf-Doped ZrTe$_{3}$

We carry out temperature-dependent scanning tunneling microscopy (STM) studies of the charge density wave (CDW) compound ZrTe$_3$ which is intentionally doped with Hf. Previous bulk studies tie Hf doping to an enhancement of the CDW transition temperature (T$_{CDW}$). In our work, by combining STM measurements with density functional theory (DFT) calculations, we observe and identify multiple defects in Zr$_{0.95}$Hf$_{0.05}$Te$_3$. Surprisingly, instead of finding clear structural or electronic signatures associated with Hf dopants, we determine the origin of the observed defects are consistent with Te and Zr vacancies. Further, our temperature dependent STM measurements allow us to examine CDW pinning to both types of observed defects below and above T$_{CDW}$.

cond-mat.mtrl-sci↗
arXiv2025

An Open-Source Framework for Quality-Assured Smartphone-Based Visible Light Iris Recognition

Smartphone-based iris recognition in the visible spectrum (VIS) offers a low-cost and accessible biometric alternative but remains a challenge due to lighting variability, pigmentation effects, and the limited adoption of standardized capture protocols. In this work, we present CUVIRIS, a dataset of 752 ISO/IEC 29794-6 compliant iris images from 47 subjects, collected with a custom Android application that enforces real-time framing, sharpness assessment, and quality feedback. We further introduce LightIrisNet, a MobileNetV3-based multi-task segmentation model optimized for on-device deployment. In addition, we adapt IrisFormer, a transformer-based matcher, to the VIS domain. We evaluate OSIRIS and IrisFormer under a standardized protocol and benchmark against published CNN baselines reported in prior work. On CUVIRIS, the open-source OSIRIS system achieves a TAR of 97.9% at FAR = 0.01 (EER = 0.76%), while IrisFormer, trained only on the UBIRIS.v2 dataset, achieves an EER of 0.057\%. To support reproducibility, we release the Android application, LightIrisNet, trained IrisFormer weights, and a subset of the CUVIRIS dataset. These results show that, with standardized acquisition and VIS-adapted lightweight models, accurate iris recognition on commodity smartphones is feasible under controlled conditions, bringing this modality closer to practical deployment.

eess.IV↗
arXiv2025

Cosmological Constraints on the Phenomenological Interacting Dark Energy Model with Fermi Gamma-Ray Bursts and DESI DR2

In this work, we constrain the phenomenological interacting dark energy (IDE) model using \emph{Fermi} gamma-ray burst (GRB) dataset and the latest baryon acoustic oscillation (BAO) data from the Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2). Through a joint Bayesian analysis, we perform a cosmological comparative assessment of the $Λ$CDM, $w$CDM, and CPL models with the phenomenological IDE model. For the phenomenological IDE model in a flat universe with \emph{Fermi} samples and DESI DR2, we obtain: $ξ=2.63^{+0.63}_{-0.52}$, $ξ+ 3w_X = -0.98^{+1.90}_{-2.07}$ with the GOLD sample ($1.4\le z \le5.6$) and $ξ=2.83^{+0.63}_{-0.58}$, $ξ+ 3w_X = 0.03^{+1.35}_{-1.33}$ with the FULL sample ($1.4\le z \le8.2$), respectively. Our analysis shows that the $Λ$CDM model without interaction ($ξ=3$, $ξ+ 3w_X = 0$) is consistent with the latest \emph{Fermi} sample and DESI DR2 at $1σ$ confidence level. We find no significant deviations from the standard model using AIC and BIC criterias.

astro-ph.CO↗
arXiv2025

FACM: Flow-Anchored Consistency Models

Continuous-time Consistency Models (CMs) promise efficient few-step generation but face significant challenges with training instability. We argue this instability stems from a fundamental conflict: Training the network exclusively on a shortcut objective leads to the catastrophic forgetting of the instantaneous velocity field that defines the flow. Our solution is to explicitly anchor the model in the underlying flow, ensuring high trajectory fidelity during training. We introduce the Flow-Anchored Consistency Model (FACM), where a Flow Matching (FM) task serves as a dynamic anchor for the primary CM shortcut objective. Key to this Flow-Anchoring approach is a novel expanded time interval strategy that unifies optimization for a single model while decoupling the two tasks to ensure stable, architecturally-agnostic training. By distilling a pre-trained LightningDiT model, our method achieves a state-of-the-art FID of 1.32 with two steps (NFE=2) and 1.70 with just one step (NFE=1) on ImageNet 256x256. To address the challenge of scalability, we develop a memory-efficient Chain-JVP that resolves key incompatibilities with FSDP. This method allows us to scale FACM training on a 14B parameter model (Wan 2.2), accelerating its Text-to-Image inference from 2x40 to 2-8 steps. Our code and pretrained models: https://github.com/ali-vilab/FACM.

cs.CV↗
arXiv2025

Anchoring Values in Temporal and Group Dimensions for Flow Matching Model Alignment

Group Relative Policy Optimization (GRPO) has proven highly effective in enhancing the alignment capabilities of Large Language Models (LLMs). However, current adaptations of GRPO for the flow matching-based image generation neglect a foundational conflict between its core principles and the distinct dynamics of the visual synthesis process. This mismatch leads to two key limitations: (i) Uniformly applying a sparse terminal reward across all timesteps impairs temporal credit assignment, ignoring the differing criticality of generation phases from early structure formation to late-stage tuning. (ii) Exclusive reliance on relative, intra-group rewards causes the optimization signal to fade as training converges, leading to the optimization stagnation when reward diversity is entirely depleted. To address these limitations, we propose Value-Anchored Group Policy Optimization (VGPO), a framework that redefines value estimation across both temporal and group dimensions. Specifically, VGPO transforms the sparse terminal reward into dense, process-aware value estimates, enabling precise credit assignment by modeling the expected cumulative reward at each generative stage. Furthermore, VGPO replaces standard group normalization with a novel process enhanced by absolute values to maintain a stable optimization signal even as reward diversity declines. Extensive experiments on three benchmarks demonstrate that VGPO achieves state-of-the-art image quality while simultaneously improving task-specific accuracy, effectively mitigating reward hacking. Project webpage: https://yawen-shao.github.io/VGPO/.

cs.LG↗
arXiv2025

How AI Agents Follow the Herd of AI? Network Effects, History, and Machine Optimism

Understanding decision-making in multi-AI-agent frameworks is crucial for analyzing strategic interactions in network-effect-driven contexts. This study investigates how AI agents navigate network-effect games, where individual payoffs depend on peer participatio--a context underexplored in multi-agent systems despite its real-world prevalence. We introduce a novel workflow design using large language model (LLM)-based agents in repeated decision-making scenarios, systematically manipulating price trajectories (fixed, ascending, descending, random) and network-effect strength. Our key findings include: First, without historical data, agents fail to infer equilibrium. Second, ordered historical sequences (e.g., escalating prices) enable partial convergence under weak network effects but strong effects trigger persistent "AI optimism"--agents overestimate participation despite contradictory evidence. Third, randomized history disrupts convergence entirely, demonstrating that temporal coherence in data shapes LLMs' reasoning, unlike humans. These results highlight a paradigm shift: in AI-mediated systems, equilibrium outcomes depend not just on incentives, but on how history is curated, which is impossible for human.

cs.MA↗
arXiv2025

Electric-Field-Controlled Altermagnetic Transition for Neuromorphic Computing

Altermagnets represent a novel magnetic phase with transformative potential for ultrafast spintronics, yet efficient control of their magnetic states remains challenging. We demonstrate an ultra-low-power electric-field control of altermagnetism in MnTe through strain-mediated coupling in MnTe/PMN-PT heterostructures with negligible Joule heating. Application of +6 kV/cm electric fields induces piezoelectric strain in PMN-PT, modulating the Néel temperature from 310 to 328 K. As a result, around the magnetic phase transition, the altermagnetic spin splitting of MnTe is reversibly switched "on" and "off" by the electric fields. Meanwhile, the piezoelectric strain generates lattice distortions and magnetic structure changes in MnTe, enabling up to 9.7% resistance modulation around the magnetic phase transition temperature. Leveraging this effect, we implement programmable resistance states in a Hopfield neuromorphic network, achieving 100% pattern recognition accuracy at <=40% noise levels. This approach establishes the electric-field control as a low-power strategy for altermagnetic manipulation while demonstrating the viability of altermagnetic materials for energy-efficient neuromorphic computing beyond conventional charge-based architectures.

cond-mat.mtrl-sci↗
arXiv2025

PathMind: A Retrieve-Prioritize-Reason Framework for Knowledge Graph Reasoning with Large Language Models

Knowledge graph reasoning (KGR) is the task of inferring new knowledge by performing logical deductions on knowledge graphs. Recently, large language models (LLMs) have demonstrated remarkable performance in complex reasoning tasks. Despite promising success, current LLM-based KGR methods still face two critical limitations. First, existing methods often extract reasoning paths indiscriminately, without assessing their different importance, which may introduce irrelevant noise that misleads LLMs. Second, while many methods leverage LLMs to dynamically explore potential reasoning paths, they require high retrieval demands and frequent LLM calls. To address these limitations, we propose PathMind, a novel framework designed to enhance faithful and interpretable reasoning by selectively guiding LLMs with important reasoning paths. Specifically, PathMind follows a "Retrieve-Prioritize-Reason" paradigm. First, it retrieves a query subgraph from KG through the retrieval module. Next, it introduces a path prioritization mechanism that identifies important reasoning paths using a semantic-aware path priority function, which simultaneously considers the accumulative cost and the estimated future cost for reaching the target. Finally, PathMind generates accurate and logically consistent responses via a dual-phase training strategy, including task-specific instruction tuning and path-wise preference alignment. Extensive experiments on benchmark datasets demonstrate that PathMind consistently outperforms competitive baselines, particularly on complex reasoning tasks with fewer input tokens, by identifying essential reasoning paths.

cs.AI↗