arXiv subjects

Han Wu

Publications and source records attributed to Han Wu.

At least 91 records · Page 5Linked to original sources

arXiv2025

Searching for heavy millicharged particles from the atmosphere

If millicharged particles (MCPs) exist they can be created in the atmosphere when high energy cosmic rays collide with nuclei and could subsequently be detected at neutrino experiments. We extend previous work, which considered MCPs from decays of light mesons and proton bremsstrahlung, by including production from $Υ$ meson decays and the Drell-Yan process. MCPs with masses below a GeV primarily arise from proton bremsstrahlung, while heavier MCPs predominantly originate from heavy meson decays and Drell-Yan. We analyse the resulting single scatter and multiple scatter signals at SuperK and JUNO. Searches for low energy coincident signals at JUNO will be sensitive to MCPs with milli-charges up to an order of magnitude beyond current constraints for MCP masses between 2 GeV and 10 GeV.

hep-ph↗

arXiv2025

Determine-Then-Ensemble: Necessity of Top-k Union for Large Language Model Ensembling

Large language models (LLMs) exhibit varying strengths and weaknesses across different tasks, prompting recent studies to explore the benefits of ensembling models to leverage their complementary advantages. However, existing LLM ensembling methods often overlook model compatibility and struggle with inefficient alignment of probabilities across the entire vocabulary. In this study, we empirically investigate the factors influencing ensemble performance, identifying model performance, vocabulary size, and response style as key determinants, revealing that compatibility among models is essential for effective ensembling. This analysis leads to the development of a simple yet effective model selection strategy that identifies compatible models. Additionally, we introduce the \textsc{Uni}on \textsc{T}op-$k$ \textsc{E}nsembling (\textsc{UniTE}), a novel approach that efficiently combines models by focusing on the union of the top-k tokens from each model, thereby avoiding the need for full vocabulary alignment and reducing computational overhead. Extensive evaluations across multiple benchmarks demonstrate that \textsc{UniTE} significantly enhances performance compared to existing methods, offering a more efficient framework for LLM ensembling.

cs.CL↗

arXiv2025

Sens-Merging: Sensitivity-Guided Parameter Balancing for Merging Large Language Models

Recent advances in large language models have led to numerous task-specialized fine-tuned variants, creating a need for efficient model merging techniques that preserve specialized capabilities while avoiding costly retraining. While existing task vector-based merging methods show promise, they typically apply uniform coefficients across all parameters, overlooking varying parameter importance both within and across tasks. We present Sens-Merging, a sensitivity-guided coefficient adjustment method that enhances existing model merging techniques by operating at both task-specific and cross-task levels. Our method analyzes parameter sensitivity within individual tasks and evaluates cross-task transferability to determine optimal merging coefficients. Extensive experiments on Mistral 7B and LLaMA2-7B/13B models demonstrate that Sens-Merging significantly improves performance across general knowledge, mathematical reasoning, and code generation tasks. Notably, when combined with existing merging techniques, our method enables merged models to outperform specialized fine-tuned models, particularly in code generation tasks. Our findings reveal important trade-offs between task-specific and cross-task scalings, providing insights for future model merging strategies.

cs.CL↗

arXiv2025

VT-GAN: Cooperative Tabular Data Synthesis using Vertical Federated Learning

This paper presents the application of Vertical Federated Learning (VFL) to generate synthetic tabular data using Generative Adversarial Networks (GANs). VFL is a collaborative approach to train machine learning models among distinct tabular data holders, such as financial institutions, who possess disjoint features for the same group of customers. In this paper we introduce the VT-GAN framework, Vertical federated Tabular GAN, and demonstrate that VFL can be successfully used to implement GANs for distributed tabular data in privacy-preserving manner, with performance close to centralized GANs that assume shared data. We make design choices with respect to the distribution of GAN generator and discriminator models and introduce a training-with-shuffling technique so that no party can reconstruct training data from the GAN conditional vector. The paper presents (1) an implementation of VT-GAN, (2) a detailed quality evaluation of the VT-GAN-generated synthetic data, (3) an overall scalability examination of VT-GAN framework, (4) a security analysis on VT-GAN's robustness against Membership Inference Attack with different settings of Differential Privacy, for a range of datasets with diverse distribution characteristics. Our results demonstrate that VT-GAN can consistently generate high-fidelity synthetic tabular data of comparable quality to that generated by a centralized GAN algorithm. The difference in machine learning utility can be as low as 2.7%, even under extremely imbalanced data distributions across clients or with different numbers of clients.

cs.LG↗

arXiv2024

Molly: Making Large Language Model Agents Solve Python Problem More Logically

Applying large language models (LLMs) as teaching assists has attracted much attention as an integral part of intelligent education, particularly in computing courses. To reduce the gap between the LLMs and the computer programming education expert, fine-tuning and retrieval augmented generation (RAG) are the two mainstream methods in existing researches. However, fine-tuning for specific tasks is resource-intensive and may diminish the model`s generalization capabilities. RAG can perform well on reducing the illusion of LLMs, but the generation of irrelevant factual content during reasoning can cause significant confusion for learners. To address these problems, we introduce the Molly agent, focusing on solving the proposed problem encountered by learners when learning Python programming language. Our agent automatically parse the learners' questioning intent through a scenario-based interaction, enabling precise retrieval of relevant documents from the constructed knowledge base. At generation stage, the agent reflect on the generated responses to ensure that they not only align with factual content but also effectively answer the user's queries. Extensive experimentation on a constructed Chinese Python QA dataset shows the effectiveness of the Molly agent, indicating an enhancement in its performance for providing useful responses to Python questions.

cs.CL↗

arXiv2024

Joint Communication and Channel Discrimination

We consider a basic joint communication and sensing setup comprising a transmitter, a receiver and a sensor. The transmitter sends a codeword to the receiver through a discrete memoryless channel, and the receiver is interested in decoding the transmitted codeword. At the same time, the sensor picks up a noisy version of the transmitted codeword through one of two possible discrete memoryless channels. The sensor knows the codeword and wishes to discriminate between the two possible channels, i.e. to identify the channel that has generated the output given the input. We study the trade-off between communication and sensing in the asymptotic regime, captured in terms of the channel coding rate against the two types of discrimination error exponents. We characterize the optimal trade-off between the rate and the exponents for general discrete memoryless channels with an input cost constraint.

cs.IT↗

arXiv2024

A privacy-preserving distributed credible evidence fusion algorithm for collective decision-making

The theory of evidence reasoning has been applied to collective decision-making in recent years. However, existing distributed evidence fusion methods lead to participants' preference leakage and fusion failures as they directly exchange raw evidence and do not assess evidence credibility like centralized credible evidence fusion (CCEF) does. To do so, a privacy-preserving distributed credible evidence fusion method with three-level consensus (PCEF) is proposed in this paper. In evidence difference measure (EDM) neighbor consensus, an evidence-free equivalent expression of EDM among neighbored agents is derived with the shared dot product protocol for pignistic probability and the identical judgment of two events with maximal subjective probabilities, so that evidence privacy is guaranteed due to such irreversible evidence transformation. In EDM network consensus, the non-neighbored EDMs are inferred and neighbored EDMs reach uniformity via interaction between linear average consensus (LAC) and low-rank matrix completion with rank adaptation to guarantee EDM consensus convergence and no solution of inferring raw evidence in numerical iteration style. In fusion network consensus, a privacy-preserving LAC with a self-cancelling differential privacy term is proposed, where each agent adds its randomness to the sharing content and step-by-step cancels such randomness in consensus iterations. Besides, the sufficient condition of the convergence to the CCEF is explored, and it is proven that raw evidence is impossibly inferred in such an iterative consensus. The simulations show that PCEF is close to CCEF both in credibility and fusion results and obtains higher decision accuracy with less time-comsuming than existing methods.

cs.AI↗

arXiv2024

MMTE: Corpus and Metrics for Evaluating Machine Translation Quality of Metaphorical Language

Machine Translation (MT) has developed rapidly since the release of Large Language Models and current MT evaluation is performed through comparison with reference human translations or by predicting quality scores from human-labeled data. However, these mainstream evaluation methods mainly focus on fluency and factual reliability, whilst paying little attention to figurative quality. In this paper, we investigate the figurative quality of MT and propose a set of human evaluation metrics focused on the translation of figurative language. We additionally present a multilingual parallel metaphor corpus generated by post-editing. Our evaluation protocol is designed to estimate four aspects of MT: Metaphorical Equivalence, Emotion, Authenticity, and Quality. In doing so, we observe that translations of figurative expressions display different traits from literal ones.

cs.CL↗

arXiv2024

Qini Curves for Multi-Armed Treatment Rules

Qini curves have emerged as an attractive and popular approach for evaluating the benefit of data-driven targeting rules for treatment allocation. We propose a generalization of the Qini curve to multiple costly treatment arms, that quantifies the value of optimally selecting among both units and treatment arms at different budget levels. We develop an efficient algorithm for computing these curves and propose bootstrap-based confidence intervals that are exact in large samples for any point on the curve. These confidence intervals can be used to conduct hypothesis tests comparing the value of treatment targeting using an optimal combination of arms with using just a subset of arms, or with a non-targeting assignment rule ignoring covariates, at different budget levels. We demonstrate the statistical performance in a simulation experiment and an application to treatment targeting for election turnout.

stat.ME↗

arXiv2024

CoCA: Regaining Safety-awareness of Multimodal Large Language Models with Constitutional Calibration

The deployment of multimodal large language models (MLLMs) has demonstrated remarkable success in engaging in conversations involving visual inputs, thanks to the superior power of large language models (LLMs). Those MLLMs are typically built based on the LLMs, with an image encoder to process images into the token embedding space of the LLMs. However, the integration of visual modality has introduced a unique vulnerability: the MLLM becomes susceptible to malicious visual inputs and prone to generating sensitive or harmful responses, even though the LLM has been trained on textual dataset to align with human value. In this paper, we first raise the question: ``Do the MLLMs possess safety-awareness against malicious image inputs?". We find that after adding a principle that specifies the safety requirement into the input of the MLLM, the model's safety awareness becomes boosted. This phenomenon verifies the existence of MLLM's safety-awareness against image inputs, it is only weakened by the modality gap. We then introduce a simple yet effective technique termed CoCA, which amplifies the safety-awareness of the MLLM by calibrating its output distribution. Our proposed strategy helps the model reclaim its original safety awareness without losing its original capabilities. We verify the effectiveness of our approach on both multimodal safety and understanding benchmarks.

cs.CL↗

arXiv2024

Persistent flat band splitting and strong selective band renormalization in a kagome magnet thin film

Magnetic kagome materials provide a fascinating playground for exploring the interplay of magnetism, correlation and topology. Many magnetic kagome systems have been reported including the binary FemXn (X=Sn, Ge; m:n = 3:1, 3:2, 1:1) family and the rare earth RMn6Sn6 (R = rare earth) family, where their kagome flat bands are calculated to be near the Fermi level in the paramagnetic phase. While partially filling a kagome flat band is predicted to give rise to a Stoner-type ferromagnetism, experimental visualization of the magnetic splitting across the ordering temperature has not been reported for any of these systems due to the high ordering temperatures, hence leaving the nature of magnetism in kagome magnets an open question. Here, we probe the electronic structure with angle-resolved photoemission spectroscopy in a kagome magnet thin film FeSn synthesized using molecular beam epitaxy. We identify the exchange-split kagome flat bands, whose splitting persists above the magnetic ordering temperature, indicative of a local moment picture. Such local moments in the presence of the topological flat band are consistent with the compact molecular orbitals predicted in theory. We further observe a large spin-orbital selective band renormalization in the Fe d_xy+d_(x^2-y^2 ) spin majority channel reminiscent of the orbital selective correlation effects in the iron-based superconductors. Our discovery of the coexistence of local moments with topological flat bands in a kagome system echoes similar findings in magic-angle twisted bilayer graphene, and provides a basis for theoretical effort towards modeling correlation effects in magnetic flat band systems.

cond-mat.mes-hall↗

arXiv2024

Automatic deductive coding in discourse analysis: an application of large language models in learning analytics

Deductive coding is a common discourse analysis method widely used by learning science and learning analytics researchers for understanding teaching and learning interactions. It often requires researchers to manually label all discourses to be analyzed according to a theoretically guided coding scheme, which is time-consuming and labor-intensive. The emergence of large language models such as GPT has opened a new avenue for automatic deductive coding to overcome the limitations of traditional deductive coding. To evaluate the usefulness of large language models in automatic deductive coding, we employed three different classification methods driven by different artificial intelligence technologies, including the traditional text classification method with text feature engineering, BERT-like pretrained language model and GPT-like pretrained large language model (LLM). We applied these methods to two different datasets and explored the potential of GPT and prompt engineering in automatic deductive coding. By analyzing and comparing the accuracy and Kappa values of these three classification methods, we found that GPT with prompt engineering outperformed the other two methods on both datasets with limited number of training samples. By providing detailed prompt structures, the reported work demonstrated how large language models can be used in the implementation of automatic deductive coding.

cs.CL↗

arXiv2024

Geometry-Aware Attenuation Learning for Sparse-View CBCT Reconstruction

Cone Beam Computed Tomography (CBCT) plays a vital role in clinical imaging. Traditional methods typically require hundreds of 2D X-ray projections to reconstruct a high-quality 3D CBCT image, leading to considerable radiation exposure. This has led to a growing interest in sparse-view CBCT reconstruction to reduce radiation doses. While recent advances, including deep learning and neural rendering algorithms, have made strides in this area, these methods either produce unsatisfactory results or suffer from time inefficiency of individual optimization. In this paper, we introduce a novel geometry-aware encoder-decoder framework to solve this problem. Our framework starts by encoding multi-view 2D features from various 2D X-ray projections with a 2D CNN encoder. Leveraging the geometry of CBCT scanning, it then back-projects the multi-view 2D features into the 3D space to formulate a comprehensive volumetric feature map, followed by a 3D CNN decoder to recover 3D CBCT image. Importantly, our approach respects the geometric relationship between 3D CBCT image and its 2D X-ray projections during feature back projection stage, and enjoys the prior knowledge learned from the data population. This ensures its adaptability in dealing with extremly sparse view inputs without individual training, such as scenarios with only 5 or 10 X-ray projections. Extensive evaluations on two simulated datasets and one real-world dataset demonstrate exceptional reconstruction quality and time efficiency of our method.

eess.IV↗

arXiv2024

Iterated descent obstructions for algebraic stacks

Base on a conjecture, we prove that for any smooth separated stack of finite type over a number field, its descent obstruction equals its iterated descent obstruction. As a consequence, we show that for any algebraic stack over a number field that has a finite etale covering of a smooth geometrically integral variety, its descent obstruction equals its iterated descent obstruction.

math.NT↗

arXiv2024

Topological Surface State Evolution in Bi$_2$Se$_3$ via Surface Etching

Topological insulators are materials with an insulating bulk interior while maintaining gapless boundary states against back scattering. Bi$_2$Se$_3$ is a prototypical topological insulator with a Dirac-cone surface state around $Γ$. Here, we present a controlled methodology to gradually remove Se atoms from the surface Se-Bi-Se-Bi-Se quintuple layers, eventually forming bilayer-Bi on top of the quintuple bulk. Our method allows us to track the topological surface state and confirm its robustness throughout the surface modification. Importantly, we report a relocation of the topological Dirac cone in both real space and momentum space, as the top surface layer transitions from quintuple Se-Bi-Se-Bi-Se to bilayer-Bi. Additionally, charge transfer among different surface layers is identified. Our study provides a precise method to manipulate surface configurations, allowing for the fine-tuning of the topological surface states in Bi$_2$Se$_3$, which represents a significant advancement towards nano-engineering of topological states.

cond-mat.mtrl-sci↗

arXiv2024

Multi-watt long-wavelength infrared femtosecond lasers and resonant enamel ablation

High-power broadband tunable long-wavelength infrared (LWIR) femtosecond lasers operating at fingerprint wavelengths of 7-14 μm hold significant promise across a range of applications, including molecular hyperspectral imaging, strong-field light-matter interaction, and resonant tissue ablation. Here we present 6-12 μm broadband tunable parametric amplifier based on LiGaS2 or BaGa4S7, generating new record output power of 2.4 W at 7.5 μm, and 1.5 W at 9.5 μm, pumped by a simple and effective thin-square-rod Yb:YAG amplifier producing 110 W 274 fs output pulses. As a proof of concept, we showcase efficient resonant ablation and microstructure fabrication on enamel at the hydroxyapatite resonant wavelength of 9.5 μm, with a laser intensity two orders-of-magnitude lower than that required by non-resonant femtosecond lasers, which could foster more precision surgical applications with superior biosafety.

physics.optics↗

arXiv2024

Learning From Correctness Without Prompting Makes LLM Efficient Reasoner

Large language models (LLMs) have demonstrated outstanding performance across various tasks, yet they still exhibit limitations such as hallucination, unfaithful reasoning, and toxic content. One potential approach to mitigate these issues is learning from human or external feedback (e.g. tools). In this paper, we introduce an intrinsic self-correct reasoning framework for LLMs that eliminates the need for human feedback, external tools, and handcraft prompts. The proposed framework, based on a multi-step reasoning paradigm \textbf{Le}arning from \textbf{Co}rrectness (\textsc{LeCo}), improves reasoning performance without needing to learn from errors. This paradigm prioritizes learning from correct reasoning steps, and a unique method to measure confidence for each reasoning step based on generation logits. Experimental results across various multi-step reasoning tasks demonstrate the effectiveness of the framework in improving reasoning performance with reduced token consumption.

cs.CL↗

arXiv2024

A Human-in-the-Middle Attack against Object Detection Systems

Object detection systems using deep learning models have become increasingly popular in robotics thanks to the rising power of CPUs and GPUs in embedded systems. However, these models are susceptible to adversarial attacks. While some attacks are limited by strict assumptions on access to the detection system, we propose a novel hardware attack inspired by Man-in-the-Middle attacks in cryptography. This attack generates a Universal Adversarial Perturbations (UAP) and injects the perturbation between the USB camera and the detection system via a hardware attack. Besides, prior research is misled by an evaluation metric that measures the model accuracy rather than the attack performance. In combination with our proposed evaluation metrics, we significantly increased the strength of adversarial perturbations. These findings raise serious concerns for applications of deep learning models in safety-critical systems, such as autonomous driving.

cs.RO↗