arXiv subjects

Hao Sun

Publications and source records attributed to Hao Sun.

At least 253 records · Page 14Linked to original sources

arXiv2025

Conservation-informed Graph Learning for Spatiotemporal Dynamics Prediction

Data-centric methods have shown great potential in understanding and predicting spatiotemporal dynamics, enabling better design and control of the object system. However, deep learning models often lack interpretability, fail to obey intrinsic physics, and struggle to cope with the various domains. While geometry-based methods, e.g., graph neural networks (GNNs), have been proposed to further tackle these challenges, they still need to find the implicit physical laws from large datasets and rely excessively on rich labeled data. In this paper, we herein introduce the conservation-informed GNN (CiGNN), an end-to-end explainable learning framework, to learn spatiotemporal dynamics based on limited training data. The network is designed to conform to the general conservation law via symmetry, where conservative and non-conservative information passes over a multiscale space enhanced by a latent temporal marching strategy. The efficacy of our model has been verified in various spatiotemporal systems based on synthetic and real-world datasets, showing superiority over baseline models. Results demonstrate that CiGNN exhibits remarkable accuracy and generalizability, and is readily applicable to learning for prediction of various spatiotemporal dynamics in a spatial domain with complex geometry.

cs.LG↗

arXiv2024

PA-RAG: RAG Alignment via Multi-Perspective Preference Optimization

The emergence of Retrieval-augmented generation (RAG) has alleviated the issues of outdated and hallucinatory content in the generation of large language models (LLMs), yet it still reveals numerous limitations. When a general-purpose LLM serves as the RAG generator, it often suffers from inadequate response informativeness, response robustness, and citation quality. Past approaches to tackle these limitations, either by incorporating additional steps beyond generating responses or optimizing the generator through supervised fine-tuning (SFT), still failed to align with the RAG requirement thoroughly. Consequently, optimizing the RAG generator from multiple preference perspectives while maintaining its end-to-end LLM form remains a challenge. To bridge this gap, we propose Multiple Perspective Preference Alignment for Retrieval-Augmented Generation (PA-RAG), a method for optimizing the generator of RAG systems to align with RAG requirements comprehensively. Specifically, we construct high-quality instruction fine-tuning data and multi-perspective preference data by sampling varied quality responses from the generator across different prompt documents quality scenarios. Subsequently, we optimize the generator using SFT and Direct Preference Optimization (DPO). Extensive experiments conducted on four question-answer datasets across three LLMs demonstrate that PA-RAG can significantly enhance the performance of RAG generators. Our code and datasets are available at https://github.com/wujwyi/PA-RAG.

cs.CL↗

arXiv2024

Few-shot Steerable Alignment: Adapting Rewards and LLM Policies with Neural Processes

As large language models (LLMs) become increasingly embedded in everyday applications, ensuring their alignment with the diverse preferences of individual users has become a critical challenge. Currently deployed approaches typically assume homogeneous user objectives and rely on single-objective fine-tuning. However, human preferences are inherently heterogeneous, influenced by various unobservable factors, leading to conflicting signals in preference data. Existing solutions addressing this diversity often require costly datasets labelled for specific objectives and involve training multiple reward models or LLM policies, which is computationally expensive and impractical. In this work, we present a novel framework for few-shot steerable alignment, where users' underlying preferences are inferred from a small sample of their choices. To achieve this, we extend the Bradley-Terry-Luce model to handle heterogeneous preferences with unobserved variability factors and propose its practical implementation for reward modelling and LLM fine-tuning. Thanks to our proposed approach of functional parameter-space conditioning, LLMs trained with our framework can be adapted to individual preferences at inference time, generating outputs over a continuum of behavioural modes. We empirically validate the effectiveness of methods, demonstrating their ability to capture and align with diverse human preferences in a data-efficient manner. Our code is made available at: https://github.com/kasia-kobalczyk/few-shot-steerable-alignment.

cs.LG↗

arXiv2024

Orbital Magnetic Moment Dynamics and Hanle Magnetoresistance in Multilayered 2D Materials

The orbital Hall effect (OHE), resulting from non-trivial quantum geometry of 2D materials, has several potential advantages over the spin Hall effect (SHE), the latter being well known for its many applications in spintronics. Like the spin Hall effect, the OHE occurs in nonmagnetic materials without stringent symmetry requirements, but unlike the SHE it does no rely on relatively weak spin-orbit interaction. In 2D materials, these advantages risk to be nullified by the difficulty of turning the orbital moment away from the out-of-plane direction. Multilayered 2D materials offer a way out of this difficulty because the fluctuating in-plane component of the orbital moment, due to motion of electrons between the layers, can latch to a magnetic field. To describe this effect we have derived a semi-phenomenological equation of motion for the density of orbital magnetic moment in stacked 2D materials subjected to a magnetic field. Unlike the equations of motion for the spin, these equations produce a strongly anisotropic dynamics, which is governed by an inverse effective mass tensor for which we provide a fully microscopic expression. As a first application, we combine our equation of motion with phenomenological drift-diffusion equations to obtain a theory of orbital Hanle magnetoresistance in multilayered 2D materials.

cond-mat.mes-hall↗

arXiv2024

Trusted Unified Feature-Neighborhood Dynamics for Multi-View Classification

Multi-view classification (MVC) faces inherent challenges due to domain gaps and inconsistencies across different views, often resulting in uncertainties during the fusion process. While Evidential Deep Learning (EDL) has been effective in addressing view uncertainty, existing methods predominantly rely on the Dempster-Shafer combination rule, which is sensitive to conflicting evidence and often neglects the critical role of neighborhood structures within multi-view data. To address these limitations, we propose a Trusted Unified Feature-NEighborhood Dynamics (TUNED) model for robust MVC. This method effectively integrates local and global feature-neighborhood (F-N) structures for robust decision-making. Specifically, we begin by extracting local F-N structures within each view. To further mitigate potential uncertainties and conflicts in multi-view fusion, we employ a selective Markov random field that adaptively manages cross-view neighborhood dependencies. Additionally, we employ a shared parameterized evidence extractor that learns global consensus conditioned on local F-N structures, thereby enhancing the global integration of multi-view features. Experiments on benchmark datasets show that our method improves accuracy and robustness over existing approaches, particularly in scenarios with high uncertainty and conflicting views. The code will be made available at https://github.com/JethroJames/TUNED.

cs.CV↗

arXiv2024

Constructing optimal treatment length strategies to maximize quality-adjusted lifetimes

Real-world clinical decision making is a complex process that involves balancing the risks and benefits of treatments. Quality-adjusted lifetime is a composite outcome that combines patient quantity and quality of life, making it an attractive outcome in clinical research. We propose methods for constructing optimal treatment length strategies to maximize this outcome. Existing methods for estimating optimal treatment strategies for survival outcomes cannot be applied to a quality-adjusted lifetime due to induced informative censoring. We propose a weighted estimating equation that adjusts for both confounding and informative censoring. We also propose a nonparametric estimator of the mean counterfactual quality-adjusted lifetime survival curve under a given treatment length strategy, where the weights are estimated using an undersmoothed sieve-based estimator. We show that the estimator is asymptotically linear and provide a data-dependent undersmoothing criterion. We apply our method to obtain the optimal time for percutaneous endoscopic gastrostomy insertion in patients with amyotrophic lateral sclerosis.

math.ST↗

arXiv2024

Artifact Correction in Magnetic Resonance Temperature Imaging for Laser Interstitial Thermotherapy with Multi-echo Acquisitions

In MRI-guided laser interstitial thermotherapy (MRgLITT), a signal void sometimes appears at the heating center of the measured temperature map. In neurosurgical MRgLITT treatments, cerebrospinal fluid pulsation (CSF), which may lead to temperature artifacts, also needs to be carefully managed. We find that signal loss in MR magnitude images can be one distinct contributor to the temperature imaging signal void. Therefore, this study aims to investigate this finding and more importantly. Also, this study intends to improve measurement accuracy by correcting CSF-induced temperature errors and employing a more reliable phase unwrapping algorithm. A gradient echo sequence with certain TE values for temperature imaging is used to quantify T2* variations during MRgLITT and to investigate the development of signal voids throughout the treatment. Informed by these findings, a multi-echo GRE sequence with appropriate TE coverage is employed. A multi-echo-based correction algorithm is developed to address the signal loss-induced temperature errors. A new phase unwrapping method and a new CSF pulsation correction approach are developed for multi-echo signal processing. The temperature imaging method is evaluated by gel phantom, ex-vivo, and in-vivo LITT heating experiments. T2* shortening during heating can be one important cause of the temperate imaging signal voids and this demands the multi-echo acquisition with varied TE values. The proposed multi-echo-based method can effectively correct signal loss-induced temperature errors and raise temperature estimation precision. The multi-echo thermometry in the in-vivo experiments shows smoother hotspot boundaries, fewer artifacts, and improved thermometry reliability. In the in-vivo experiments, the ablation areas estimated from the multi-echo thermometry also show satisfactory agreement with those determined from post-ablation MR imaging.

physics.med-ph↗

arXiv2024

Accelerating Non-Maximum Suppression: A Graph Theory Perspective

Non-maximum suppression (NMS) is an indispensable post-processing step in object detection. With the continuous optimization of network models, NMS has become the ``last mile'' to enhance the efficiency of object detection. This paper systematically analyzes NMS from a graph theory perspective for the first time, revealing its intrinsic structure. Consequently, we propose two optimization methods, namely QSI-NMS and BOE-NMS. The former is a fast recursive divide-and-conquer algorithm with negligible mAP loss, and its extended version (eQSI-NMS) achieves optimal complexity of $\mathcal{O}(n\log n)$. The latter, concentrating on the locality of NMS, achieves an optimization at a constant level without an mAP loss penalty. Moreover, to facilitate rapid evaluation of NMS methods for researchers, we introduce NMS-Bench, the first benchmark designed to comprehensively assess various NMS methods. Taking the YOLOv8-N model on MS COCO 2017 as the benchmark setup, our method QSI-NMS provides $6.2\times$ speed of original NMS on the benchmark, with a $0.1\%$ decrease in mAP. The optimal eQSI-NMS, with only a $0.3\%$ mAP decrease, achieves $10.7\times$ speed. Meanwhile, BOE-NMS exhibits $5.1\times$ speed with no compromise in mAP.

cs.CV↗

arXiv2024

Rich Semantic Knowledge Enhanced Large Language Models for Few-shot Chinese Spell Checking

Chinese Spell Checking (CSC) is a widely used technology, which plays a vital role in speech to text (STT) and optical character recognition (OCR). Most of the existing CSC approaches relying on BERT architecture achieve excellent performance. However, limited by the scale of the foundation model, BERT-based method does not work well in few-shot scenarios, showing certain limitations in practical applications. In this paper, we explore using an in-context learning method named RS-LLM (Rich Semantic based LLMs) to introduce large language models (LLMs) as the foundation model. Besides, we study the impact of introducing various Chinese rich semantic information in our framework. We found that by introducing a small number of specific Chinese rich semantic structures, LLMs achieve better performance than the BERT-based model on few-shot CSC task. Furthermore, we conduct experiments on multiple datasets, and the experimental results verified the superiority of our proposed framework.

cs.CL↗

arXiv2024

Z_5 two-component dark matter in the Type-II seesaw mechanism

We consider the Z5 two-component dark matter model within the framework of the Type-II seesaw mechanism. Due to the new annihilation processes related to triplets, the light component cannot necessarily be dominant in the dark matter relic density, which is different from the Z5 two-component dark matter model in the SM. The model is considered to explain the excess of electron-positron flux measured by the AMS-02 Collaborations in this work, which is encouraged by the decay of the triplets arising from dark matter annihilations in the Galactic halo. We discuss the cases of the light and heavy components determining dark matter density within a viable parameter space satisfying relic density and direct detection constraints, and by fitting the antiproton spectrum observed in the PAMELA and AMS experiments, we find that the parameter space is flexible and the electron-positron flux excess can be obtained in both cases with the mass of two dark matter particles being larger than that of the triplets'.

hep-ph↗

arXiv2024

Nonconserved Density Accumulations in Orbital Hall Transport: Insights from Linear Response Theory

We present a linear response theory for stationary density accumulations in anomalous transport phenomena, such as the orbital Hall effect, where the transported density is odd under time reversal and the underlying charge is not conserved. Our framework applies to both metals and insulators, topologically trivial or nontrivial, and distinguishes between contributions from bulk and edge states, as well as undergap and dissipative currents. In time-reversal invariant systems, we prove a microscopic reciprocity theorem showing that only dissipative currents at the Fermi level contribute to density accumulation, while undergap currents do not. In contrast, in non-time-reversal invariant systems, non-dissipative density accumulations, such as magnetoelectric polarization, can appear in both the bulk and edges. Importantly, we find that the net density accumulation does not always vanish, pointing to a global non-conservation that implies the existence of a non-vanishing integrated ``net torque'' in addition to a ``distributed torque'', which has zero spatial average. We show that the distributed torque can be absorbed in the divergence of a redefined current that satisfies Onsager reciprocity, while the net torque must be explicitly accounted for. Finally, we apply our theory to two-dimensional models with edge terminations.

cond-mat.mes-hall↗

arXiv2024

StreamAdapter: Efficient Test Time Adaptation from Contextual Streams

In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks directly from the given demonstrations without requiring gradient updates. While recent advances have expanded context windows to accommodate more demonstrations, this approach increases inference costs without necessarily improving performance. To mitigate these issues, We propose StreamAdapter, a novel approach that directly updates model parameters from context at test time, eliminating the need for explicit in-context demonstrations. StreamAdapter employs context mapping and weight absorption mechanisms to dynamically transform ICL demonstrations into parameter updates with minimal additional parameters. By reducing reliance on numerous in-context examples, StreamAdapter significantly reduce inference costs and allows for efficient inference with constant time complexity, regardless of demonstration count. Extensive experiments across diverse tasks and model architectures demonstrate that StreamAdapter achieves comparable or superior adaptation capability to ICL while requiring significantly fewer demonstrations. The superior task adaptation and context encoding capabilities of StreamAdapter on both language understanding and generation tasks provides a new perspective for adapting LLMs at test time using context, allowing for more efficient adaptation across scenarios and more cost-effective inference

cs.CL↗

arXiv2024

Statistical Operating Characteristics of Current Early Phase Dose Finding Designs with Toxicity and Efficacy in Oncology

Traditional phase I dose finding cancer clinical trial designs aim to determine the maximum tolerated dose (MTD) of the investigational cytotoxic agent based on a single toxicity outcome, assuming a monotone dose-response relationship. However, this assumption might not always hold for newly emerging therapies such as immuno-oncology therapies and molecularly targeted therapies, making conventional dose finding trial designs based on toxicity no longer appropriate. To tackle this issue, numerous early phase dose finding clinical trial designs have been developed to identify the optimal biological dose (OBD), which takes both toxicity and efficacy outcomes into account. In this article, we review the current model-assisted dose finding designs, BOIN-ET, BOIN12, UBI, TEPI-2, PRINTE, STEIN, and uTPI to identify the OBD and compare their operating characteristics. Extensive simulation studies and a case study using a CAR T-cell therapy phase I trial have been conducted to compare the performance of the aforementioned designs under different possible dose-response relationship scenarios. The simulation results demonstrate that the performance of different designs varies depending on the particular dose-response relationship and the specific metric considered. Based on our simulation results and practical considerations, STEIN, PRINTE, and BOIN12 outperform the other designs from different perspectives.

stat.ME↗

arXiv2024

The Theoretical Study of $p\bar{p}\to\barΛΣη$ Reaction

We study the production of hyperon resonances in the $p\bar{p}\to\barΛΣη$ reaction within an effective Lagrangian approach. The model includes the production of $Σ(1750)$ and $Λ(1670)$ in the intermediate state excited by the $K$ and $K^*$ meson exchanges between the initial proton and antiproton. Due to the large coupling of $Σ(1750)Ση$ vertex, $Σ(1750)$ is found a significant contribution near the threshold in this reaction. We provide total and differential cross section predictions for the reation and discuss the possible influence of $Σ(1750)Ση$ vertex coupling and model parameters, which will be useful in future experimental studies. This reaction can provide a platform for studying the features of $Σ(1750)$ resonance, especially the coupling to $Ση$ channel.

hep-ph↗

arXiv2024

Over-parameterized Student Model via Tensor Decomposition Boosted Knowledge Distillation

Increased training parameters have enabled large pre-trained models to excel in various downstream tasks. Nevertheless, the extensive computational requirements associated with these models hinder their widespread adoption within the community. We focus on Knowledge Distillation (KD), where a compact student model is trained to mimic a larger teacher model, facilitating the transfer of knowledge of large models. In contrast to much of the previous work, we scale up the parameters of the student model during training, to benefit from overparameterization without increasing the inference latency. In particular, we propose a tensor decomposition strategy that effectively over-parameterizes the relatively small student model through an efficient and nearly lossless decomposition of its parameter matrices into higher-dimensional tensors. To ensure efficiency, we further introduce a tensor constraint loss to align the high-dimensional tensors between the student and teacher models. Comprehensive experiments validate the significant performance enhancement by our approach in various KD tasks, covering computer vision and natural language processing areas. Our code is available at https://github.com/intell-sci-comput/OPDF.

cs.AI↗

arXiv2024

ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors

The safety of Large Language Models (LLMs) has gained increasing attention in recent years, but there still lacks a comprehensive approach for detecting safety issues within LLMs' responses in an aligned, customizable and explainable manner. In this paper, we propose ShieldLM, an LLM-based safety detector, which aligns with common safety standards, supports customizable detection rules, and provides explanations for its decisions. To train ShieldLM, we compile a large bilingual dataset comprising 14,387 query-response pairs, annotating the safety of responses based on various safety standards. Through extensive experiments, we demonstrate that ShieldLM surpasses strong baselines across four test sets, showcasing remarkable customizability and explainability. Besides performing well on standard detection datasets, ShieldLM has also been shown to be effective as a safety evaluator for advanced LLMs. ShieldLM is released at \url{https://github.com/thu-coai/ShieldLM} to support accurate and explainable safety detection under various safety standards.

cs.CL↗

arXiv2024

Token-level Proximal Policy Optimization for Query Generation

Query generation is a critical task for web search engines (e.g. Google, Bing) and recommendation systems. Recently, state-of-the-art query generation methods leverage Large Language Models (LLMs) for their strong capabilities in context understanding and text generation. However, they still face challenges in generating high-quality queries in terms of inferring user intent based on their web search interaction history. In this paper, we propose Token-level Proximal Policy Optimization (TPPO), a noval approach designed to empower LLMs perform better in query generation through fine-tuning. TPPO is based on the Reinforcement Learning from AI Feedback (RLAIF) paradigm, consisting of a token-level reward model and a token-level proximal policy optimization module to address the sparse reward challenge in traditional RLAIF frameworks. To evaluate the effectiveness and robustness of TPPO, we conducted experiments on both open-source dataset and an industrial dataset that was collected from a globally-used search engine. The experimental results demonstrate that TPPO significantly improves the performance of query generation for LLMs and outperforms its existing competitors.

cs.LG↗

arXiv2024

SUANPAN: Scalable Photonic Linear Vector Machine

Photonic linear operation is a promising approach to handle the extensive vector multiplications in artificial intelligence techniques due to the natural bosonic parallelism and high-speed information transmission of photonics. Although it is believed that maximizing the interaction of the light beams is necessary to fully utilize the parallelism and tremendous efforts have been made in past decades, the achieved dimensionality of vector-matrix multiplication is very limited due to the difficulty of scaling up a tightly interconnected or highly coupled optical system. Additionally, there is still a lack of a universal photonic computing architecture that can be readily merged with existing computing system to meet the computing power demand of AI techniques. Here, we propose a programmable and reconfigurable photonic linear vector machine to perform only the inner product of two vectors, formed by a series of independent basic computing units, while each unit is just one pair of light-emitter and photodetector. Since there is no interaction among light beams inside, extreme scalability could be achieved by simply duplicating the independent basic computing unit while there is no requirement of large-scale analog-to-digital converter and digital-to-analog converter arrays. Our architecture is inspired by the traditional Chinese Suanpan or abacus and thus is denoted as photonic SUANPAN. As a proof of principle, SUANPAN architecture is implemented with an 8*8 vertical cavity surface emission laser array and an 8*8 MoTe2 two-dimensional material photodetector array. We believe that our proposed photonic SUANPAN is capable of serving as a fundamental linear vector machine that can be readily merged with existing electronic digital computing system and is potential to enhance the computing power for future various AI applications.

physics.optics↗