arXiv subjects

Yu Wang

Publications and source records attributed to Yu Wang.

At least 865 records · Page 48Linked to original sources

arXiv2023

Classical Shadow Tomography with Mutually Unbiased Bases

Classical shadow tomography, harnessing randomized informationally complete (IC) measurements, provides an effective avenue for predicting many properties of unknown quantum states with sample-efficient precision. Projections onto $2^n+1$ mutually unbiased bases (MUBs) are widely recognized as minimal and optimal IC measurements for full-state tomography. We study how to use MUBs circuits as the ensemble in classical shadow tomography. For the general observables, the variance to predict their expectation value is shown to be exponential to the number of qubits $n$. However, for a special class termed as appropriate MUBs-average (AMA) observables, the variance decreases to $poly(n)$. Additionally, we find that through biased sampling of MUBs circuits, the variance for non-AMA observables can again be reduced to $poly(n)$ with the MUBs-sparse condition. The performance and complexity of using the MUBs and Clifford circuits as the ensemble in the classical shadow tomography are compared in the end.

quant-ph↗

arXiv2023

DeSCo: Towards Generalizable and Scalable Deep Subgraph Counting

We introduce DeSCo, a scalable neural deep subgraph counting pipeline, designed to accurately predict both the count and occurrence position of queries on target graphs post single training. Firstly, DeSCo uses a novel canonical partition and divides the large target graph into small neighborhood graphs, greatly reducing the count variation while guaranteeing no missing or double-counting. Secondly, neighborhood counting uses an expressive subgraph-based heterogeneous graph neural network to accurately count in each neighborhood. Finally, gossip propagation propagates neighborhood counts with learnable gates to harness the inductive biases of motif counts. DeSCo is evaluated on eight real-world datasets from various domains. It outperforms state-of-the-art neural methods with 137x improvement in the mean squared error of count prediction, while maintaining the polynomial runtime complexity. Our open source project is at https://github.com/fuvty/DeSCo.

cs.LG↗

arXiv2023

Label Denoising through Cross-Model Agreement

Learning from corrupted labels is very common in real-world machine-learning applications. Memorizing such noisy labels could affect the learning of the model, leading to sub-optimal performances. In this work, we propose a novel framework to learn robust machine-learning models from noisy labels. Through an empirical study, we find that different models make relatively similar predictions on clean examples, while the predictions on noisy examples vary much more across different models. Motivated by this observation, we propose \em denoising with cross-model agreement \em (DeCA) which aims to minimize the KL-divergence between the true label distributions parameterized by two machine learning models while maximizing the likelihood of data observation. We employ the proposed DeCA on both the binary label scenario and the multiple label scenario. For the binary label scenario, we select implicit feedback recommendation as the downstream task and conduct experiments with four state-of-the-art recommendation models on four datasets. For the multiple-label scenario, the downstream application is image classification on two benchmark datasets. Experimental results demonstrate that the proposed methods significantly improve the model performance compared with normal training and other denoising methods on both binary and multiple-label scenarios.

cs.LG↗

arXiv2023

EncryIP: A Practical Encryption-Based Framework for Model Intellectual Property Protection

In the rapidly growing digital economy, protecting intellectual property (IP) associated with digital products has become increasingly important. Within this context, machine learning (ML) models, being highly valuable digital assets, have gained significant attention for IP protection. This paper introduces a practical encryption-based framework called \textit{EncryIP}, which seamlessly integrates a public-key encryption scheme into the model learning process. This approach enables the protected model to generate randomized and confused labels, ensuring that only individuals with accurate secret keys, signifying authorized users, can decrypt and reveal authentic labels. Importantly, the proposed framework not only facilitates the protected model to multiple authorized users without requiring repetitive training of the original ML model with IP protection methods but also maintains the model's performance without compromising its accuracy. Compared to existing methods like watermark-based, trigger-based, and passport-based approaches, \textit{EncryIP} demonstrates superior effectiveness in both training protected models and efficiently detecting the unauthorized spread of ML models.

cs.CR↗

arXiv2023

DRDT: Dynamic Reflection with Divergent Thinking for LLM-based Sequential Recommendation

The rise of Large Language Models (LLMs) has sparked interest in their application to sequential recommendation tasks as they can provide supportive item information. However, due to the inherent complexities of sequential recommendation, such as sequential patterns across datasets, noise within sequences, and the temporal evolution of user preferences, existing LLM reasoning strategies, such as in-context learning and chain-of-thought are not fully effective. To address these challenges, we introduce a novel reasoning principle: Dynamic Reflection with Divergent Thinking within a retriever-reranker framework. Our approach starts with a collaborative in-context demonstration retriever, which collects sequences exhibiting collaborative behaviors as in-context examples. Following this, we abstract high-level user preferences across multiple aspects, providing a more nuanced understanding of user interests and circumventing the noise within the raw sequences. The cornerstone of our methodology is dynamic reflection, a process that emulates human learning through probing, critiquing, and reflecting, using user feedback to tailor the analysis more effectively to the target user in a temporal manner. We evaluate our approach on three datasets using six pre-trained LLMs. The superior performance observed across these models demonstrates the efficacy of our reasoning strategy, notably achieved without the need to fine-tune the LLMs. With our principle, we managed to outperform GPT-Turbo-3.5 on three datasets using 7b models e.g., Vicuna-7b and Openchat-7b on NDCG@10. This research not only highlights the potential of LLMs in enhancing sequential recommendation systems but also underscores the importance of developing tailored reasoning strategies to fully harness their capabilities.

cs.IR↗

arXiv2023

Deciphering Compatibility Relationships with Textual Descriptions via Extraction and Explanation

Understanding and accurately explaining compatibility relationships between fashion items is a challenging problem in the burgeoning domain of AI-driven outfit recommendations. Present models, while making strides in this area, still occasionally fall short, offering explanations that can be elementary and repetitive. This work aims to address these shortcomings by introducing the Pair Fashion Explanation (PFE) dataset, a unique resource that has been curated to illuminate these compatibility relationships. Furthermore, we propose an innovative two-stage pipeline model that leverages this dataset. This fine-tuning allows the model to generate explanations that convey the compatibility relationships between items. Our experiments showcase the model's potential in crafting descriptions that are knowledgeable, aligned with ground-truth matching correlations, and that produce understandable and informative descriptions, as assessed by both automatic metrics and human evaluation. Our code and data are released at https://github.com/wangyu-ustc/PairFashionExplanation

cs.CL↗

arXiv2023

Accelerate Multi-Agent Reinforcement Learning in Zero-Sum Games with Subgame Curriculum Learning

Learning Nash equilibrium (NE) in complex zero-sum games with multi-agent reinforcement learning (MARL) can be extremely computationally expensive. Curriculum learning is an effective way to accelerate learning, but an under-explored dimension for generating a curriculum is the difficulty-to-learn of the subgames -- games induced by starting from a specific state. In this work, we present a novel subgame curriculum learning framework for zero-sum games. It adopts an adaptive initial state distribution by resetting agents to some previously visited states where they can quickly learn to improve performance. Building upon this framework, we derive a subgame selection metric that approximates the squared distance to NE values and further adopt a particle-based state sampler for subgame generation. Integrating these techniques leads to our new algorithm, Subgame Automatic Curriculum Learning (SACL), which is a realization of the subgame curriculum learning framework. SACL can be combined with any MARL algorithm such as MAPPO. Experiments in the particle-world environment and Google Research Football environment show SACL produces much stronger policies than baselines. In the challenging hide-and-seek quadrant environment, SACL produces all four emergent stages and uses only half the samples of MAPPO with self-play. The project website is at https://sites.google.com/view/sacl-rl.

cs.LG↗

arXiv2023

Learning Language-guided Adaptive Hyper-modality Representation for Multimodal Sentiment Analysis

Though Multimodal Sentiment Analysis (MSA) proves effective by utilizing rich information from multiple sources (e.g., language, video, and audio), the potential sentiment-irrelevant and conflicting information across modalities may hinder the performance from being further improved. To alleviate this, we present Adaptive Language-guided Multimodal Transformer (ALMT), which incorporates an Adaptive Hyper-modality Learning (AHL) module to learn an irrelevance/conflict-suppressing representation from visual and audio features under the guidance of language features at different scales. With the obtained hyper-modality representation, the model can obtain a complementary and joint representation through multimodal fusion for effective MSA. In practice, ALMT achieves state-of-the-art performance on several popular datasets (e.g., MOSI, MOSEI and CH-SIMS) and an abundance of ablation demonstrates the validity and necessity of our irrelevance/conflict suppression mechanism.

cs.AI↗

arXiv2023

A Unified Sampling Framework for Solver Searching of Diffusion Probabilistic Models

Recent years have witnessed the rapid progress and broad application of diffusion probabilistic models (DPMs). Sampling from DPMs can be viewed as solving an ordinary differential equation (ODE). Despite the promising performance, the generation of DPMs usually consumes much time due to the large number of function evaluations (NFE). Though recent works have accelerated the sampling to around 20 steps with high-order solvers, the sample quality with less than 10 NFE can still be improved. In this paper, we propose a unified sampling framework (USF) to study the optional strategies for solver. Under this framework, we further reveal that taking different solving strategies at different timesteps may help further decrease the truncation error, and a carefully designed \emph{solver schedule} has the potential to improve the sample quality by a large margin. Therefore, we propose a new sampling framework based on the exponential integral formulation that allows free choices of solver strategy at each step and design specific decisions for the framework. Moreover, we propose $S^3$, a predictor-based search method that automatically optimizes the solver schedule to get a better time-quality trade-off of sampling. We demonstrate that $S^3$ can find outstanding solver schedules which outperform the state-of-the-art sampling methods on CIFAR-10, CelebA, ImageNet, and LSUN-Bedroom datasets. Specifically, we achieve 2.69 FID with 10 NFE and 6.86 FID with 5 NFE on CIFAR-10 dataset, outperforming the SOTA method significantly. We further apply $S^3$ to Stable-Diffusion model and get an acceleration ratio of 2$\times$, showing the feasibility of sampling in very few steps without retraining the neural network.

cs.AI↗

arXiv2023

Spatial Knowledge-Infused Hierarchical Learning: An Application in Flood Mapping on Earth Imagery

Deep learning for Earth imagery plays an increasingly important role in geoscience applications such as agriculture, ecology, and natural disaster management. Still, progress is often hindered by the limited training labels. Given Earth imagery with limited training labels, a base deep neural network model, and a spatial knowledge base with label constraints, our problem is to infer the full labels while training the neural network. The problem is challenging due to the sparse and noisy input labels, spatial uncertainty within the label inference process, and high computational costs associated with a large number of sample locations. Existing works on neuro-symbolic models focus on integrating symbolic logic into neural networks (e.g., loss function, model architecture, and training label augmentation), but these methods do not fully address the challenges of spatial data (e.g., spatial uncertainty, the trade-off between spatial granularity and computational costs). To bridge this gap, we propose a novel Spatial Knowledge-Infused Hierarchical Learning (SKI-HL) framework that iteratively infers sample labels within a multi-resolution hierarchy. Our framework consists of a module to selectively infer labels in different resolutions based on spatial uncertainty and a module to train neural network parameters with uncertainty-aware multi-instance learning. Extensive experiments on real-world flood mapping datasets show that the proposed model outperforms several baseline methods. The code is available at \url{https://github.com/ZelinXu2000/SKI-HL}.

cs.AI↗

arXiv2023

Electric Field Switching of Magnon Spin Current in a Compensated Ferrimagnet

Manipulation of directional magnon propagation, known as magnon spin current, is essential for developing magnonic memory and logic devices featuring nonvolatile functionalities and ultralow power consumption. Magnon spin current can usually be modulated by magnetic field or current-induced spin torques. However, these approaches may lead to energy dissipation caused by Joule heating. Electric-field switching of magnon spin current without charge current is highly desired but very challenging to realize. By integrating magnonic and piezoelectric materials, we demonstrate manipulation of the magnon spin current generated by the spin Seebeck effect in the ferrimagnetic insulator Gd3Fe5O12 (GdIG) film on a piezoelectric substrate. We observe reversible electric-field switching of magnon polarization without applied charge current. Through strain-mediated magnetoelectric coupling, the electric field induces the magnetic compensation transition between two magnetic states of the GdIG, resulting in its magnetization reversal and the simultaneous switching of magnon spin current. Our work establishes a prototype material platform that pave the way for developing magnon logic devices characterized by all electric field reading and writing and reveals the underlying physics principles of their functions.

cond-mat.mes-hall↗

arXiv2023

Adaptive Tuning of Robotic Polishing Skills based on Force Feedback Model

Acquiring human skills offers an efficient approach to tackle complex task planning challenges. When performing a learned skill model for a continuous contact task, such as robot polishing in an uncertain environment, the robot needs to be able to adaptively modify the skill model to suit the environment and perform the desired task. The environmental perturbation of the polishing task is mainly reflected in the variation of contact force. Therefore, adjusting the task skill model by providing feedback on the contact force deviation is an effective way to meet the task requirements. In this study, a phase-modulated diagonal recurrent neural network (PMDRNN) is proposed for force feedback model learning in the robotic polishing task. The contact between the tool and the workpiece in the polishing task can be considered a dynamic system. In comparison to the existing feedforward neural network phase-modulated neural network (PMNN), PMDRNN combines the diagonal recurrent network structure with the phase-modulated neural network layer to improve the learning performance of the feedback model for dynamic systems. Specifically, data from real-world robot polishing experiments are used to learn the feedback model. PMDRNN demonstrates a significant reduction in the training error of the feedback model when compared to PMNN. Building upon this, the combination of PMDRNN and dynamic movement primitives (DMPs) can be used for real-time adjustment of skills for polishing tasks and effectively improve the robustness of the task skill model. Finally, real-world robotic polishing experiments are conducted to demonstrate the effectiveness of the approach.

cs.RO↗

arXiv2023

Prospect of detecting magnetic fields from strong-magnetized binary neutron stars

Binary neutron star mergers are unique sources of gravitational waves in multi-messenger astronomy. The inspiral phase of binary neutron stars can emit gravitational waves as chirp signals. The present waveform models of gravitational wave only considered the gravitational interaction. In this paper, we derive the waveform of the gravitational wave signal taking into account the presence of magnetic fields. We found that the electromagnetic interaction and radiation can introduce different frequency-dependent power laws for both amplitude and frequency of the gravitational wave. We show from the results of Fisher information matrix that the third-generation observation may detect magnetic dipole moments if the magnetic field is around 10^17 G.

gr-qc↗

arXiv2023

Measuring the Hubble constant with coalescences of binary neutron star and neutron star-black hole: bright sirens \& dark sirens

The observations of gravitational wave (GW) provide us a new probe to study the universe. GW events can be used as standard sirens if their redshifts are measured. Normally, stardard sirens can be divided into bright/dark sirens according to whether the redshifts are measured by electromagnetic (EM) counterpart observations. Firstly, we investigate the capability of the 2.5-meter Wide-Field Survey Telescope (WFST) to take follow-up observations of kilonova counterparts. For binary neutron star (BNS) bright sirens, WFST is expected to observe 10-20 kilonovae per year in the second-generation (2G) GW detection era. As for neutron star-black hole (NSBH) mergers, when a BH spin is extremely high and the NS is stiff, the observation rate is $\sim10$ per year. Combining optical and GW observations, the bright sirens are expected to constrain the Hubble constant $H_0$ to $\sim2.8\%$ in five years of observations. As for dark sirens, tidal effects of neutron stars (NSs) during merging time provide us a cosmological model-independent approach to measure the redshifts of GW sources. Then we investigate the applications of tidal effects in redshift measurements. We find in 3G era, the host galaxy groups of around 45\% BNS mergers at $z<0.1$ can be identified through this method, if the EOS is ms1, which is roughly equivalent to the results from luminosity distant constraints. Therefore, tidal effect observations provide a reliable and cosmological model-independent method of identifying BNS mergers' host galaxy groups. Using this method, the BNS/NSBH dark sirens can constrain $H_0$ to 0.2\%/0.3\% over a five-year observation period.

astro-ph.HE↗

arXiv2023

Observation of zero-energy modes in Gd atomic chains on superconducting Nb(110)

In this experimental study, we use scanning tunneling microscopy and spectroscopy to investigate Yu-Shiba-Rusinov states induced by 4f-shell rare-earth Gd adatoms on a superconducting Nb(110) surface. We engineer Gd atom chains along the substrate's $[1\overline{1}0]$ and $[001]$ directions, revealing distinct behaviors in differently oriented chains. $[1\overline{1}0]$-oriented Gd chains exhibit spectroscopic features at their ends, identifying them as trivial edge states, while $[001]$-oriented Gd chains display zero-energy edge states, suggesting non-trivial nature. Notably, Gd chains with four atoms--independent of their particular orientation--exhibit a uniform zero-energy mode along the entire chain. These findings call for further research and a theoretical framework to describe rare-earth-based structures on superconductors.

cond-mat.supr-con↗

arXiv2023

MedEval: A Multi-Level, Multi-Task, and Multi-Domain Medical Benchmark for Language Model Evaluation

Curated datasets for healthcare are often limited due to the need of human annotations from experts. In this paper, we present MedEval, a multi-level, multi-task, and multi-domain medical benchmark to facilitate the development of language models for healthcare. MedEval is comprehensive and consists of data from several healthcare systems and spans 35 human body regions from 8 examination modalities. With 22,779 collected sentences and 21,228 reports, we provide expert annotations at multiple levels, offering a granular potential usage of the data and supporting a wide range of tasks. Moreover, we systematically evaluated 10 generic and domain-specific language models under zero-shot and finetuning settings, from domain-adapted baselines in healthcare to general-purposed state-of-the-art large language models (e.g., ChatGPT). Our evaluations reveal varying effectiveness of the two categories of language models across different tasks, from which we notice the importance of instruction tuning for few-shot usage of large language models. Our investigation paves the way toward benchmarking language models for healthcare and provides valuable insights into the strengths and limitations of adopting large language models in medical domains, informing their practical applications and future advancements.

cs.CL↗

arXiv2023

Creation of Two-Mode Squeezed States in Atomic Mechanical Oscillators

Two-mode squeezed states, which are entangled states with bipartite quantum correlations in continuous-variable systems, are crucial in quantum information processing and metrology. Recently, continuous-variable quantum computing with the vibrational modes of trapped atoms has emerged with significant progress, featuring a high degree of control in hybridizing with spin qubits. Creating two-mode squeezed states in such a platform could enable applications that are only viable with photons. Here, we experimentally demonstrate two-mode squeezed states by employing atoms in a two-dimensional optical lattice as quantum registers. The states are generated by a controlled projection conditioned on the relative phase of two independent squeezed states. The individual squeezing is created by sudden jumps of the oscillators' frequencies, allowing generating of the two-mode squeezed states at a rate within a fraction of the oscillation frequency. We validate the states by entanglement steering criteria and Fock state analysis. Our results can be applied in other mechanical oscillators for quantum sensing and continuous-variable quantum information.

quant-ph↗

arXiv2023

PepLand: a large-scale pre-trained peptide representation model for a comprehensive landscape of both canonical and non-canonical amino acids

In recent years, the scientific community has become increasingly interested on peptides with non-canonical amino acids due to their superior stability and resistance to proteolytic degradation. These peptides present promising modifications to biological, pharmacological, and physiochemical attributes in both endogenous and engineered peptides. Notwithstanding their considerable advantages, the scientific community exhibits a conspicuous absence of an effective pre-trained model adept at distilling feature representations from such complex peptide sequences. We herein propose PepLand, a novel pre-training architecture for representation and property analysis of peptides spanning both canonical and non-canonical amino acids. In essence, PepLand leverages a comprehensive multi-view heterogeneous graph neural network tailored to unveil the subtle structural representations of peptides. Empirical validations underscore PepLand's effectiveness across an array of peptide property predictions, encompassing protein-protein interactions, permeability, solubility, and synthesizability. The rigorous evaluation confirms PepLand's unparalleled capability in capturing salient synthetic peptide features, thereby laying a robust foundation for transformative advances in peptide-centric research domains. We have made all the source code utilized in this study publicly accessible via GitHub at https://github.com/zhangruochi/pepland

q-bio.BM↗