arXiv subjects

Yang Zhao

Publications and source records attributed to Yang Zhao.

At least 145 records · Page 8Linked to original sources

arXiv2025

Digital Gatekeepers: Exploring Large Language Model's Role in Immigration Decisions

With globalization and increasing immigrant populations, immigration departments face significant work-loads and the challenge of ensuring fairness in decision-making processes. Integrating artificial intelligence offers a promising solution to these challenges. This study investigates the potential of large language models (LLMs),such as GPT-3.5 and GPT-4, in supporting immigration decision-making. Utilizing a mixed-methods approach,this paper conducted discrete choice experiments and in-depth interviews to study LLM decision-making strategies and whether they are fair. Our findings demonstrate that LLMs can align their decision-making with human strategies, emphasizing utility maximization and procedural fairness. Meanwhile, this paper also reveals that while ChatGPT has safeguards to prevent unintentional discrimination, it still exhibits stereotypes and biases concerning nationality and shows preferences toward privileged group. This dual analysis highlights both the potential and limitations of LLMs in automating and enhancing immigration decisions.

cs.CL↗

arXiv2025

Multimodal Inconsistency Reasoning (MMIR): A New Benchmark for Multimodal Reasoning Models

Existing Multimodal Large Language Models (MLLMs) are predominantly trained and tested on consistent visual-textual inputs, leaving open the question of whether they can handle inconsistencies in real-world, layout-rich content. To bridge this gap, we propose the Multimodal Inconsistency Reasoning (MMIR) benchmark to assess MLLMs' ability to detect and reason about semantic mismatches in artifacts such as webpages, presentation slides, and posters. MMIR comprises 534 challenging samples, each containing synthetically injected errors across five reasoning-heavy categories: Factual Contradiction, Identity Misattribution, Contextual Mismatch, Quantitative Discrepancy, and Temporal/Spatial Incoherence. We evaluate six state-of-the-art MLLMs, showing that models with dedicated multimodal reasoning capabilities, such as o1, substantially outperform their counterparts while open-source models remain particularly vulnerable to inconsistency errors. Detailed error analyses further show that models excel in detecting pairwise inconsistencies but struggle with inconsistencies confined to single elements in complex layouts. Probing experiments reveal that single-modality prompting, including Chain-of-Thought (CoT) and Set-of-Mark (SoM) methods, yields marginal gains, revealing a key bottleneck in cross-modal reasoning. Our findings highlight the need for advanced multimodal reasoning and point to future research on multimodal inconsistency.

cs.CL↗

arXiv2025

Probing Millikelvin Temperature Sensitivity in Chiral Nanoparticles via Optical Forces

With increasing interest in utilizing nanostructures as nanoscale heat sources, the ability to precisely measure photothermal effects at the nanoscale has become increasingly significant. Techniques based on fluorescence or Raman signals often suffer from challenges in accurate calibration, far-field imaging methods are limited by diffraction-limited spatial resolution, and electron microscopy requires vacuum conditions, restricting in situ applicability. In contrast, tip-based measurement techniques offer sub-diffraction spatial resolution under ambient conditions, making them well-suited for nanoscale photothermal mapping. In this study, we employ tip-based optical force nanoscopy combined with phase-informed decomposition to investigate the origin of the photothermal force, enable nanoscale mapping, and evaluate temperature sensitivity. Our system achieves a temperature sensitivity of approximately 0.1 K without necessitating an additional temperature-sensitive layer. We anticipate that our approach has the potential to serve as a versatile platform for investigating localized thermal effects in fields such as semiconductors, nanophotonics, and photocatalysis.

physics.optics↗

arXiv2025

StimuVAR: Spatiotemporal Stimuli-aware Video Affective Reasoning with Multimodal Large Language Models

Predicting and reasoning how a video would make a human feel is crucial for developing socially intelligent systems. Although Multimodal Large Language Models (MLLMs) have shown impressive video understanding capabilities, they tend to focus more on the semantic content of videos, often overlooking emotional stimuli. Hence, most existing MLLMs fall short in estimating viewers' emotional reactions and providing plausible explanations. To address this issue, we propose StimuVAR, a spatiotemporal Stimuli-aware framework for Video Affective Reasoning (VAR) with MLLMs. StimuVAR incorporates a two-level stimuli-aware mechanism: frame-level awareness and token-level awareness. Frame-level awareness involves sampling video frames with events that are most likely to evoke viewers' emotions. Token-level awareness performs tube selection in the token space to make the MLLM concentrate on emotion-triggered spatiotemporal regions. Furthermore, we create VAR instruction data to perform affective training, steering MLLMs' reasoning strengths towards emotional focus and thereby enhancing their affective reasoning ability. To thoroughly assess the effectiveness of VAR, we provide a comprehensive evaluation protocol with extensive metrics. StimuVAR is the first MLLM-based method for viewer-centered VAR. Experiments demonstrate its superiority in understanding viewers' emotional responses to videos and providing coherent and insightful explanations. Our code is available at https://github.com/EthanG97/StimuVAR

cs.CV↗

arXiv2025

A Graph-Retrieval-Augmented Generation Framework Enhances Decision-Making in the Circular Economy

Large language models (LLMs) hold promise for sustainable manufacturing, but often hallucinate industrial codes and emission factors, undermining regulatory and investment decisions. We introduce CircuGraphRAG, a retrieval-augmented generation (RAG) framework that grounds LLMs outputs in a domain-specific knowledge graph for the circular economy. This graph connects 117,380 industrial and waste entities with classification codes and GWP100 emission data, enabling structured multi-hop reasoning. Natural language queries are translated into SPARQL and verified subgraphs are retrieved to ensure accuracy and traceability. Compared with Standalone LLMs and Naive RAG, CircuGraphRAG achieves superior performance in single-hop and multi-hop question answering, with ROUGE-L F1 scores up to 1.0, while baseline scores below 0.08. It also improves efficiency, halving the response time and reducing token usage by 16% in representative tasks. CircuGraphRAG provides fact-checked, regulatory-ready support for circular economy planning, advancing reliable, low-carbon resource decision making.

cs.AI↗

arXiv2025

Dynamic Domain Adaptation-Driven Physics-Informed Graph Representation Learning for AC-OPF

Alternating Current Optimal Power Flow (AC-OPF) aims to optimize generator power outputs by utilizing the non-linear relationships between voltage magnitudes and phase angles in a power system. However, current AC-OPF solvers struggle to effectively represent the complex relationship between variable distributions in the constraint space and their corresponding optimal solutions. This limitation in constraint modeling restricts the system's ability to develop diverse knowledge representations. Additionally, modeling the power grid solely based on spatial topology further limits the integration of additional prior knowledge, such as temporal information. To overcome these challenges, we propose DDA-PIGCN (Dynamic Domain Adaptation-Driven Physics-Informed Graph Convolutional Network), a new method designed to address constraint-related issues and build a graph-based learning framework that incorporates spatiotemporal features. DDA-PIGCN improves consistency optimization for features with varying long-range dependencies by applying multi-layer, hard physics-informed constraints. It also uses a dynamic domain adaptation learning mechanism that iteratively updates and refines key state variables under predefined constraints, enabling precise constraint verification. Moreover, it captures spatiotemporal dependencies between generators and loads by leveraging the physical structure of the power grid, allowing for deep integration of topological information across time and space. Extensive comparative and ablation studies show that DDA-PIGCN delivers strong performance across several IEEE standard test cases (such as case9, case30, and case300), achieving mean absolute errors (MAE) from 0.0011 to 0.0624 and constraint satisfaction rates between 99.6% and 100%, establishing it as a reliable and efficient AC-OPF solver.

cs.LG↗

arXiv2025

Information Gain-Guided Causal Intervention for Autonomous Debiasing Large Language Models

Despite significant progress, recent studies indicate that current large language models (LLMs) may still capture dataset biases and utilize them during inference, leading to the poor generalizability of LLMs. However, due to the diversity of dataset biases and the insufficient nature of bias suppression based on in-context learning, the effectiveness of previous prior knowledge-based debiasing methods and in-context learning based automatic debiasing methods is limited. To address these challenges, we explore the combination of causal mechanisms with information theory and propose an information gain-guided causal intervention debiasing (ICD) framework. To eliminate biases within the instruction-tuning dataset, it is essential to ensure that these biases do not provide any additional information to predict the answers, i.e., the information gain of these biases for predicting the answers needs to be 0. Under this guidance, this framework utilizes a causal intervention-based data rewriting method to automatically and autonomously balance the distribution of instruction-tuning dataset for reducing the information gain. Subsequently, it employs a standard supervised fine-tuning process to train LLMs on the debiased dataset. Experimental results show that ICD can effectively debias LLM to improve its generalizability across different tasks.

cs.CL↗

arXiv2025

Towards Reliable and Interpretable Traffic Crash Pattern Prediction and Safety Interventions Using Customized Large Language Models

Predicting crash events is crucial for understanding crash distributions and their contributing factors, thereby enabling the design of proactive traffic safety policy interventions. However, existing methods struggle to interpret the complex interplay among various sources of traffic crash data, including numeric characteristics, textual reports, crash imagery, environmental conditions, and driver behavior records. As a result, they often fail to capture the rich semantic information and intricate interrelationships embedded in these diverse data sources, limiting their ability to identify critical crash risk factors. In this research, we propose TrafficSafe, a framework that adapts LLMs to reframe crash prediction and feature attribution as text-based reasoning. A multi-modal crash dataset including 58,903 real-world reports together with belonged infrastructure, environmental, driver, and vehicle information is collected and textualized into TrafficSafe Event Dataset. By customizing and fine-tuning LLMs on this dataset, the TrafficSafe LLM achieves a 42% average improvement in F1-score over baselines. To interpret these predictions and uncover contributing factors, we introduce TrafficSafe Attribution, a sentence-level feature attribution framework enabling conditional risk analysis. Findings show that alcohol-impaired driving is the leading factor in severe crashes, with aggressive and impairment-related behaviors having nearly twice the contribution for severe crashes compared to other driver behaviors. Furthermore, TrafficSafe Attribution highlights pivotal features during model training, guiding strategic crash data collection for iterative performance improvements. The proposed TrafficSafe offers a transformative leap in traffic safety research, providing a blueprint for translating advanced AI technologies into responsible, actionable, and life-saving outcomes.

cs.CL↗

arXiv2025

DC-Scene: Data-Centric Learning for 3D Scene Understanding

3D scene understanding plays a fundamental role in vision applications such as robotics, autonomous driving, and augmented reality. However, advancing learning-based 3D scene understanding remains challenging due to two key limitations: (1) the large scale and complexity of 3D scenes lead to higher computational costs and slower training compared to 2D counterparts; and (2) high-quality annotated 3D datasets are significantly scarcer than those available for 2D vision. These challenges underscore the need for more efficient learning paradigms. In this work, we propose DC-Scene, a data-centric framework tailored for 3D scene understanding, which emphasizes enhancing data quality and training efficiency. Specifically, we introduce a CLIP-driven dual-indicator quality (DIQ) filter, combining vision-language alignment scores with caption-loss perplexity, along with a curriculum scheduler that progressively expands the training pool from the top 25% to 75% of scene-caption pairs. This strategy filters out noisy samples and significantly reduces dependence on large-scale labeled 3D data. Extensive experiments on ScanRefer and Nr3D demonstrate that DC-Scene achieves state-of-the-art performance (86.1 CIDEr with the top-75% subset vs. 85.4 with the full dataset) while reducing training cost by approximately two-thirds, confirming that a compact set of high-quality samples can outperform exhaustive training. Code will be available at https://github.com/AIGeeksGroup/DC-Scene.

cs.CV↗

arXiv2025

UFO-RL: Uncertainty-Focused Optimization for Efficient Reinforcement Learning Data Selection

Scaling RL for LLMs is computationally expensive, largely due to multi-sampling for policy optimization and evaluation, making efficient data selection crucial. Inspired by the Zone of Proximal Development (ZPD) theory, we hypothesize LLMs learn best from data within their potential comprehension zone. Addressing the limitation of conventional, computationally intensive multi-sampling methods for data assessment, we introduce UFO-RL. This novel framework uses a computationally efficient single-pass uncertainty estimation to identify informative data instances, achieving up to 185x faster data evaluation. UFO-RL leverages this metric to select data within the estimated ZPD for training. Experiments show that training with just 10% of data selected by UFO-RL yields performance comparable to or surpassing full-data training, reducing overall training time by up to 16x while enhancing stability and generalization. UFO-RL offers a practical and highly efficient strategy for scaling RL fine-tuning of LLMs by focusing learning on valuable data.

cs.LG↗

arXiv2025

Exploring Federated Unlearning: Review, Comparison, and Insights

The increasing demand for privacy-preserving machine learning has spurred interest in federated unlearning, which enables the selective removal of data from models trained in federated systems. However, developing federated unlearning methods presents challenges, particularly in balancing three often conflicting objectives: privacy, accuracy, and efficiency. This paper provides a comprehensive analysis of existing federated unlearning approaches, examining their algorithmic efficiency, impact on model accuracy, and effectiveness in preserving privacy. We discuss key trade-offs among these dimensions and highlight their implications for practical applications across various domains. Additionally, we propose the OpenFederatedUnlearning framework, a unified benchmark for evaluating federated unlearning methods, incorporating classic baselines and diverse performance metrics. Our findings aim to guide practitioners in navigating the complex interplay of these objectives, offering insights to achieve effective and efficient federated unlearning. Finally, we outline directions for future research to further advance the state of federated unlearning techniques.

cs.LG↗

arXiv2025

CodeV: Empowering LLMs with HDL Generation through Multi-Level Summarization

The design flow of processors, particularly in hardware description languages (HDL) like Verilog and Chisel, is complex and costly. While recent advances in large language models (LLMs) have significantly improved coding tasks in software languages such as Python, their application in HDL generation remains limited due to the scarcity of high-quality HDL data. Traditional methods of adapting LLMs for hardware design rely on synthetic HDL datasets, which often suffer from low quality because even advanced LLMs like GPT perform poorly in the HDL domain. Moreover, these methods focus solely on chat tasks and the Verilog language, limiting their application scenarios. In this paper, we observe that: (1) HDL code collected from the real world is of higher quality than code generated by LLMs. (2) LLMs like GPT-3.5 excel in summarizing HDL code rather than generating it. (3) An explicit language tag can help LLMs better adapt to the target language when there is insufficient data. Based on these observations, we propose an efficient LLM fine-tuning pipeline for HDL generation that integrates a multi-level summarization data synthesis process with a novel Chat-FIM-Tag supervised fine-tuning method. The pipeline enhances the generation of HDL code from natural language descriptions and enables the handling of various tasks such as chat and infilling incomplete code. Utilizing this pipeline, we introduce CodeV, a series of HDL generation LLMs. Among them, CodeV-All not only possesses a more diverse range of language abilities, i.e. Verilog and Chisel, and a broader scope of tasks, i.e. Chat and fill-in-middle (FIM), but it also achieves performance on VerilogEval that is comparable to or even surpasses that of CodeV-Verilog fine-tuned on Verilog only, making them the first series of open-source LLMs designed for multi-scenario HDL generation.

cs.PL↗

arXiv2025

Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model

This technical report presents a cost-efficient strategy for training a video generation foundation model. We present a mid-sized research model with approximately 7 billion parameters (7B) called Seaweed-7B trained from scratch using 665,000 H100 GPU hours. Despite being trained with moderate computational resources, Seaweed-7B demonstrates highly competitive performance compared to contemporary video generation models of much larger size. Design choices are especially crucial in a resource-constrained setting. This technical report highlights the key design decisions that enhance the performance of the medium-sized diffusion model. Empirically, we make two observations: (1) Seaweed-7B achieves performance comparable to, or even surpasses, larger models trained on substantially greater GPU resources, and (2) our model, which exhibits strong generalization ability, can be effectively adapted across a wide range of downstream applications either by lightweight fine-tuning or continue training. See the project page at https://seaweed.video/

cs.CV↗

arXiv2025

Enumeration of idempotent-sum subsequences in finite cyclic semigroups and smooth sequences

The enumeration of zero-sum subsequences of a given sequence over finite cyclic groups is one classical topic, which starts from one question of P. Erdős. In this paper, we consider this problem in a more general setting -- finite cyclic semigroups. Let $\mathcal{S}$ be a finite cyclic semigroup. By $\textbf{e}$ we denote the unique idempotent of the semigroup $\mathcal{S}$. Let $T$ be a sequence over the semigroup $\mathcal{S}$, and let $N(T; \textbf{e})$ be the number of distinct subsequences of $T$ with sum being the idempotent $\textbf{e}$. We obtain the lower bound for $N(T; \textbf{e})$ in terms of the length of $T$, and moreover, prove that $T$ contains subsequences with some smooth-structure in case that $N(T; \textbf{e})$ is not large. Our result generalizes the theorem obtained by W. Gao [Discrete Math., 1994] on the enumeration of zero-sum subsequences over finite cyclic groups to the setting of semigroups.

math.CO↗

arXiv2025

Photonic logic tensor computing beyond TOPS per core

The soaring demand for computing resources has spurred great interest in photonic computing with higher speed and larger computing capacity. Photonic logic gates are of crucial importance due to the fundamental role of Boolean logic in modern digital computing systems. However, most photonic logic schemes struggle to exhibit the capability of massively parallel processing and flexible reconfiguration, owing to weak and fixed nonlinearity in optical elements. Here, we propose a photonic logic tensor computing architecture for the first time and fabricate the photonic universal logic tensor core (PULTC) with a parallel logic computing capacity beyond TOPS. Ten wavelength channels and four spatial channels are designed in PULTC, where the logic computing speed in each channel can reach 50 Gbit/s. After the nonlinear mapping of microring modulators, arbitrary logic operations can be achieved by configuring the Mach-Zehnder interferometer mesh. Our work offers an innovative route for photonic universal logic computing with high-parallel capability and propels the practical applications of photonic logic computing.

physics.optics↗

arXiv2025

MediAug: Exploring Visual Augmentation in Medical Imaging

Data augmentation is essential in medical imaging for improving classification accuracy, lesion detection, and organ segmentation under limited data conditions. However, two significant challenges remain. First, a pronounced domain gap between natural photographs and medical images can distort critical disease features. Second, augmentation studies in medical imaging are fragmented and limited to single tasks or architectures, leaving the benefits of advanced mix-based strategies unclear. To address these challenges, we propose a unified evaluation framework with six mix-based augmentation methods integrated with both convolutional and transformer backbones on brain tumour MRI and eye disease fundus datasets. Our contributions are threefold. (1) We introduce MediAug, a comprehensive and reproducible benchmark for advanced data augmentation in medical imaging. (2) We systematically evaluate MixUp, YOCO, CropMix, CutMix, AugMix, and SnapMix with ResNet-50 and ViT-B backbones. (3) We demonstrate through extensive experiments that MixUp yields the greatest improvement on the brain tumor classification task for ResNet-50 with 79.19% accuracy and SnapMix yields the greatest improvement for ViT-B with 99.44% accuracy, and that YOCO yields the greatest improvement on the eye disease classification task for ResNet-50 with 91.60% accuracy and CutMix yields the greatest improvement for ViT-B with 97.94% accuracy. Code will be available at https://github.com/AIGeeksGroup/MediAug.

cs.CV↗

arXiv2025

Movable Antenna-Aided Federated Learning with Over-the-Air Aggregation: Joint Optimization of Positioning, Beamforming, and User Selection

Federated learning (FL) in wireless computing effectively utilizes communication bandwidth, yet it is vulnerable to errors during the analog aggregation process. While removing users with unfavorable channel conditions can mitigate these errors, it also reduces the available local training data for FL, which in turn hinders the convergence rate of the training process. To tackle this issue, we propose the use of movable antenna (MA) techniques to enhance the degrees of freedom within the channel space, ultimately boosting the convergence speed of FL training. Moreover, we develop a coordinated approach for uplink receiver beamforming, user selection, and MA positioning to optimize the convergence rate of wireless FL training in dynamic wireless environments. This stochastic optimization challenge is reformulated into a mixed-integer programming problem by utilizing the training loss upper bound. We then introduce a penalty dual decomposition (PDD) method to solve the mixed-integer mixed programming problem. Experimental results indicate that incorporating MA techniques significantly accelerates the training convergence of FL and greatly surpasses conventional methods.

eess.SP↗

arXiv2025

Modular Hamiltonian of holographic time band states

A holographic time band is a causal incomplete boundary spacetime subregion whose causal wedge is a causal complete bulk spacetime subregion. In an AdS$_3$ spacetime with a specifically modified IR geometry, its causal wedge coincides with its entanglement wedge, which suggests the existence of a local modular Hamiltonian for the holographic time band state. In this work, we construct the local modular Hamiltonian for holographic time bands using two independent methods: from the quantum information properties of the time band state and from the construction of consistent geometric modular flows. Both methods lead to the same unique result of the local modular Hamiltonian, reflecting the intrinsic property of the time band state. The entanglement first law has also been checked to hold for the simplest time band state. This is a substantial addition to the known holographic subsystems with a local modular Hamiltonian, beyond the few cases previously identified.

hep-th↗