arXiv subjects

Jiang Zhang

Publications and source records attributed to Jiang Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2025

Speeding up adiabatic holonomic quantum gates via $π$-pulse modulation

Holonomic quantum computation (HQC) offers an inherently robust approach to quantum gate implementation by exploiting quantum holonomies. While adiabatic HQC benefits from robustness against certain control errors, its long runtime limits practical utility due to increased exposure to environmental noise. Nonadiabatic HQC addresses this issue by enabling faster gate operations but compromises robustness. In this work, we propose a scheme for fast holonomic quantum gates based on the $π$-pulse method, which accelerates adiabatic evolution while preserving its robustness. By guiding the system Hamiltonian along geodesic paths in the parameter space and applying phase-modulating $π$ pulses at discrete points, we realize a universal set of holonomic gates beyond the conventional adiabatic limit. Our scheme allows for arbitrary single-qubit and two-qubit controlled gates within a single-loop evolution and provides additional tunable parameters for flexible gate design. These results demonstrate a promising path toward high-fidelity, fast, and robust quantum computation.

quant-ph↗

arXiv2025

Optimizing Recall or Relevance? A Multi-Task Multi-Head Approach for Item-to-Item Retrieval in Recommendation

The task of item-to-item (I2I) retrieval is to identify a set of relevant and highly engaging items based on a given trigger item. It is a crucial component in modern recommendation systems, where users' previously engaged items serve as trigger items to retrieve relevant content for future engagement. However, existing I2I retrieval models in industry are primarily built on co-engagement data and optimized using the recall measure, which overly emphasizes co-engagement patterns while failing to capture semantic relevance. This often leads to overfitting short-term co-engagement trends at the expense of long-term benefits such as discovering novel interests and promoting content diversity. To address this challenge, we propose MTMH, a Multi-Task and Multi-Head I2I retrieval model that achieves both high recall and semantic relevance. Our model consists of two key components: 1) a multi-task learning loss for formally optimizing the trade-off between recall and semantic relevance, and 2) a multi-head I2I retrieval architecture for retrieving both highly co-engaged and semantically relevant items. We evaluate MTMH using proprietary data from a commercial platform serving billions of users and demonstrate that it can improve recall by up to 14.4% and semantic relevance by up to 56.6% compared with prior state-of-the-art models. We also conduct live experiments to verify that MTMH can enhance both short-term consumption metrics and long-term user-experience-related metrics. Our work provides a principled approach for jointly optimizing I2I recall and semantic relevance, which has significant implications for improving the overall performance of recommendation systems.

cs.IR↗

arXiv2025

PCDCNet: A Surrogate Model for Air Quality Forecasting with Physical-Chemical Dynamics and Constraints

Air quality forecasting (AQF) is critical for public health and environmental management, yet remains challenging due to the complex interplay of emissions, meteorology, and chemical transformations. Traditional numerical models, such as CMAQ and WRF-Chem, provide physically grounded simulations but are computationally expensive and rely on uncertain emission inventories. Deep learning models, while computationally efficient, often struggle with generalization due to their lack of physical constraints. To bridge this gap, we propose PCDCNet, a surrogate model that integrates numerical modeling principles with deep learning. PCDCNet explicitly incorporates emissions, meteorological influences, and domain-informed constraints to model pollutant formation, transport, and dissipation. By combining graph-based spatial transport modeling, recurrent structures for temporal accumulation, and representation enhancement for local interactions, PCDCNet achieves state-of-the-art (SOTA) performance in 72-hour station-level PM2.5 and O3 forecasting while significantly reducing computational costs. Furthermore, our model is deployed in an online platform, providing free, real-time air quality forecasts, demonstrating its scalability and societal impact. By aligning deep learning with physical consistency, PCDCNet offers a practical and interpretable solution for AQF, enabling informed decision-making for both personal and regulatory applications.

cs.LG↗

arXiv2025

Preserving Privacy and Utility in LLM-Based Product Recommendations

Large Language Model (LLM)-based recommendation systems leverage powerful language models to generate personalized suggestions by processing user interactions and preferences. Unlike traditional recommendation systems that rely on structured data and collaborative filtering, LLM-based models process textual and contextual information, often using cloud-based infrastructure. This raises privacy concerns, as user data is transmitted to remote servers, increasing the risk of exposure and reducing control over personal information. To address this, we propose a hybrid privacy-preserving recommendation framework which separates sensitive from nonsensitive data and only shares the latter with the cloud to harness LLM-powered recommendations. To restore lost recommendations related to obfuscated sensitive data, we design a de-obfuscation module that reconstructs sensitive recommendations locally. Experiments on real-world e-commerce datasets show that our framework achieves almost the same recommendation utility with a system which shares all data with an LLM, while preserving privacy to a large extend. Compared to obfuscation-only techniques, our approach improves HR@10 scores and category distribution alignment, offering a better balance between privacy and recommendation quality. Furthermore, our method runs efficiently on consumer-grade hardware, making privacy-aware LLM-based recommendation systems practical for real-world use.

cs.IR↗

arXiv2025

SpinML: Customized Synthetic Data Generation for Private Training of Specialized ML Models

Specialized machine learning (ML) models tailored to users needs and requests are increasingly being deployed on smart devices with cameras, to provide personalized intelligent services taking advantage of camera data. However, two primary challenges hinder the training of such models: the lack of publicly available labeled data suitable for specialized tasks and the inaccessibility of labeled private data due to concerns about user privacy. To address these challenges, we propose a novel system SpinML, where the server generates customized Synthetic image data to Privately traIN a specialized ML model tailored to the user request, with the usage of only a few sanitized reference images from the user. SpinML offers users fine-grained, object-level control over the reference images, which allows user to trade between the privacy and utility of the generated synthetic data according to their privacy preferences. Through experiments on three specialized model training tasks, we demonstrate that our proposed system can enhance the performance of specialized models without compromising users privacy preferences.

cs.LG↗

arXiv2025

Dynamical Reversibility and A New Theory of Causal Emergence based on SVD

The theory of causal emergence (CE) with effective information (EI) posits that complex systems can exhibit CE, where macro-dynamics show stronger causal effects than micro-dynamics. A key challenge of this theory is its dependence on coarse-graining method. In this paper, we introduce a fresh concept of approximate dynamical reversibility and establish a novel framework for CE based on this. By applying singular value decomposition(SVD) to Markov dynamics, we find that the essence of CE lies in the presence of redundancy, represented by the irreversible and correlated information pathways. Therefore, CE can be quantified as the potential maximal efficiency increase for dynamical reversibility or information transmission. We also demonstrate a strong correlation between the approximate dynamical reversibility and EI, establishing an equivalence between the SVD and EI maximization frameworks for quantifying CE, supported by theoretical insights and numerical examples from Boolean networks, cellular automata, and complex networks. Importantly, our SVD-based CE framework is independent of specific coarse-graining techniques and effectively captures the fundamental characteristics of the dynamics.

cond-mat.stat-mech↗

arXiv2025

Mixing Phases and Metastability for the Glauber Dynamics on the p-Spin Curie-Weiss Model

The Glauber dynamics for the classical $2$-spin Curie-Weiss model on $N$ nodes with inverse temperature $β$ and zero external field is known to mix in time $Θ(N\log N)$ for $β< \frac{1}{2}$, in time $Θ(N^{3/2})$ at $β= \frac{1}{2}$, and in time $\exp(Ω(N))$ for $β>\frac{1}{2}$. In this paper, we consider the $p$-spin generalization of the Curie-Weiss model with an external field $h$, and identify three disjoint regions almost exhausting the parameter space, with the corresponding Glauber dynamics exhibiting three different orders of mixing times in these regions. The construction of these disjoint regions depends on the number of local maximizers of a certain function $H_{β,h,p}$, and the behavior of the second derivative of $H_{β,h,p}$ at such a local maximizer. Specifically, we show that if $H_{β,h,p}$ has a unique local maximizer $m_*$ with $H_{β,h,p}''(m_*) < 0$ and no other stationary point, then the Glauber dynamics mixes in time $Θ(N\log N)$, and if $H_{β,h,p}$ has multiple local maximizers, then the mixing time is $\exp(Ω(N))$. Finally, if $H_{β,h,p}$ has a unique local maximizer $m_*$ with $H_{β,h,p}''(m_*) = 0$, then the mixing time is $Θ(N^{3/2})$. We provide an explicit description of the geometry of these three different phases in the parameter space, and observe that the only portion of the parameter plane that is left out by the union of these three regions, is a one-dimensional curve, on which the function $H_{β,h,p}$ has a stationary inflection point. Finding out the exact order of the mixing time on this curve remains an open question. Finally, we show that if $H_{β,h,p}$ has multiple local maximizers (metastable states), then one can create a restricted version of the original Glauber dynamics, which still mixes in time $Θ(N\log N)$.

math.PR↗

arXiv2025

Multi-Omics Fusion with Soft Labeling for Enhanced Prediction of Distant Metastasis in Nasopharyngeal Carcinoma Patients after Radiotherapy

Omics fusion has emerged as a crucial preprocessing approach in the field of medical image processing, providing significant assistance to several studies. One of the challenges encountered in the integration of omics data is the presence of unpredictability arising from disparities in data sources and medical imaging equipment. In order to overcome this challenge and facilitate the integration of their joint application to specific medical objectives, this study aims to develop a fusion methodology that mitigates the disparities inherent in omics data. The utilization of the multi-kernel late-fusion method has gained significant popularity as an effective strategy for addressing this particular challenge. An efficient representation of the data may be achieved by utilizing a suitable single-kernel function to map the inherent features and afterward merging them in a space with a high number of dimensions. This approach effectively addresses the differences noted before. The inflexibility of label fitting poses a constraint on the use of multi-kernel late-fusion methods in complex nasopharyngeal carcinoma (NPC) datasets, hence affecting the efficacy of general classifiers in dealing with high-dimensional characteristics. This innovative methodology aims to increase the disparity between the two cohorts, hence providing a more flexible structure for the allocation of labels. The examination of the NPC-ContraParotid dataset demonstrates the model's robustness and efficacy, indicating its potential as a valuable tool for predicting distant metastases in patients with nasopharyngeal carcinoma (NPC).

q-bio.QM↗

arXiv2025

An Exact Theory of Causal Emergence for Linear Stochastic Iteration Systems

After coarse-graining a complex system, the dynamics of its macro-state may exhibit more pronounced causal effects than those of its micro-state. This phenomenon, known as causal emergence, is quantified by the indicator of effective information. However, two challenges confront this theory: the absence of well-developed frameworks in continuous stochastic dynamical systems and the reliance on coarse-graining methodologies. In this study, we introduce an exact theoretic framework for causal emergence within linear stochastic iteration systems featuring continuous state spaces and Gaussian noise. Building upon this foundation, we derive an analytical expression for effective information across general dynamics and identify optimal linear coarse-graining strategies that maximize the degree of causal emergence when the dimension averaged uncertainty eliminated by coarse-graining has an upper bound. Our investigation reveals that the maximal causal emergence and the optimal coarse-graining methods are primarily determined by the principal eigenvalues and eigenvectors of the dynamic system's parameter matrix, with the latter not being unique. To validate our propositions, we apply our analytical models to three simplified physical systems, comparing the outcomes with numerical simulations, and consistently achieve congruent results.

cs.IT↗

arXiv2025

Quantifying system-environment synergistic information by effective information decomposition

What is the most crucial characteristic of a system with life activity? Currently, many theories have attempted to explain the most essential difference between living systems and general systems, such as the self-organization theory and the free energy principle, but there is a lack of a reasonable indicator that can measure to what extent a system can be regarded as a system with life characteristics, especially the lack of attention to the dynamic characteristics of life systems. In this article, we propose a new indicator at the level of dynamic mechanisms to measure the ability of a system to flexibly respond to the environment. We proved that this indicator satisfies the axiom system of multivariate information decomposition in the partial information decomposition (PID) framework. Through further disassembly and analysis of this indicator, we found that it is determined by the degree of entanglement between system and environmental variables in the dynamics and the magnitude of noise. We conducted measurements on cellular automata (CA), random Boolean networks, and real gene regulatory networks (GRN), verified its relationship with the type of CA and the Langton parameter, and identified that the feedback loops have high abilities to flexibly respond to the environment on the GRN. We also combined machine learning technology to prove that this framework can be applied in the case of unknown dynamics.

q-bio.MN↗

arXiv2024

System Information Decomposition

To characterize the complex higher-order interactions among variables within a system, this study introduces a novel framework, termed System Information Decomposition (SID), aimed at decomposing the information entropy of variables into information atoms based on their interrelations. Diverging from the established Partial Information Decomposition (PID) framework, which predominantly concentrates on the directional interactions stemming from an array of source variables to a single target variable, SID adopts a holistic approach, scrutinizing the interactions across all variables within the system. Specifically, we proved all the information atoms are symmetric, which means the disentanglement of unique, redundant, and synergistic information from any specific target variable. Hence, our proposed SID framework can capture the symmetric pairwise and higher-order relationships among variables. This advance positions SID as a promising framework with the potential to foster a deeper understanding of higher-order relationships within complex systems across disciplines.

cs.IT↗

arXiv2024

Differentially Private Federated Learning without Noise Addition: When is it Possible?

Federated Learning (FL) with Secure Aggregation (SA) has gained significant attention as a privacy preserving framework for training machine learning models while preventing the server from learning information about users' data from their individual encrypted model updates. Recent research has extended privacy guarantees of FL with SA by bounding the information leakage through the aggregate model over multiple training rounds thanks to leveraging the "noise" from other users' updates. However, the privacy metric used in that work (mutual information) measures the on-average privacy leakage, without providing any privacy guarantees for worse-case scenarios. To address this, in this work we study the conditions under which FL with SA can provide worst-case differential privacy guarantees. Specifically, we formally identify the necessary condition that SA can provide DP without addition noise. We then prove that when the randomness inside the aggregated model update is Gaussian with non-singular covariance matrix, SA can provide differential privacy guarantees with the level of privacy $ε$ bounded by the reciprocal of the minimum eigenvalue of the covariance matrix. However, we further demonstrate that in practice, these conditions are almost unlikely to hold and hence additional noise added in model updates is still required in order for SA in FL to achieve DP. Lastly, we discuss the potential solution of leveraging inherent randomness inside aggregated model update to reduce the amount of addition noise required for DP guarantee.

cs.CR↗

arXiv2024

Data driven modeling for self-similar dynamics

Multiscale modeling of complex systems is crucial for understanding their intricacies. Data-driven multiscale modeling has emerged as a promising approach to tackle challenges associated with complex systems. On the other hand, self-similarity is prevalent in complex systems, hinting that large-scale complex systems can be modeled at a reduced cost. In this paper, we introduce a multiscale neural network framework that incorporates self-similarity as prior knowledge, facilitating the modeling of self-similar dynamical systems. For deterministic dynamics, our framework can discern whether the dynamics are self-similar. For uncertain dynamics, it can compare and determine which parameter set is closer to self-similarity. The framework allows us to extract scale-invariant kernels from the dynamics for modeling at any scale. Moreover, our method can identify the power law exponents in self-similar systems. Preliminary tests on the Ising model yielded critical exponents consistent with theoretical expectations, providing valuable insights for addressing critical phase transitions in non-equilibrium systems.

cs.LG↗

arXiv2024

Protecting logical qubits with dynamical decoupling

Demonstrating that logical qubits outperform their physical counterparts is a milestone for achieving reliable quantum computation. Here, we propose to protect logical qubits with a novel dynamical decoupling scheme that implements iSWAP gates on nearest-neighbor physical qubits, and experimentally demonstrate the scheme on superconducting transmon qubits. In our scheme, each logical qubit only requires two physical qubits. A universal set of quantum gates on the logical qubits can be achieved such that each logical gate comprises only one or two physical gates. Our experiments reveal that the coherence time of a logical qubit is extended by up to 366% when compared to the better-performing physical qubit. Moreover, to the best of our knowledge, we demonstrate for the first time that multiple logical qubits outperform their physical counterparts in superconducting qubits. We illustrate a set of universal gates through a logical Ramsey experiment and the creation of a logical Bell state. Given its scalable nature, our scheme holds promise as a component for future reliable quantum computation.

quant-ph↗

arXiv2024

Learning nonequilibrium statistical mechanics and dynamical phase transitions

Nonequilibrium statistical mechanics exhibit a variety of complex phenomena far from equilibrium. It inherits challenges of equilibrium, including accurately describing the joint distribution of a large number of configurations, and also poses new challenges as the distribution evolves over time. Characterizing dynamical phase transitions as an emergent behavior further requires tracking nonequilibrium systems under a control parameter. While a number of methods have been proposed, such as tensor networks for one-dimensional lattices, we lack a method for arbitrary time beyond the steady state and for higher dimensions. Here, we develop a general computational framework to study the time evolution of nonequilibrium systems in statistical mechanics by leveraging variational autoregressive networks, which offer an efficient computation on the dynamical partition function, a central quantity for discovering the phase transition. We apply the approach to prototype models of nonequilibrium statistical mechanics, including the kinetically constrained models of structural glasses up to three dimensions. The approach uncovers the active-inactive phase transition of spin flips, the dynamical phase diagram, as well as new scaling relations. The result highlights the potential of machine learning dynamical phase transitions in nonequilibrium systems.

cond-mat.stat-mech↗

arXiv2024

Location Leakage in Federated Signal Maps

We consider the problem of predicting cellular network performance (signal maps) from measurements collected by several mobile devices. We formulate the problem within the online federated learning framework: (i) federated learning (FL) enables users to collaboratively train a model, while keeping their training data on their devices; (ii) measurements are collected as users move around over time and are used for local training in an online fashion. We consider an honest-but-curious server, who observes the updates from target users participating in FL and infers their location using a deep leakage from gradients (DLG) type of attack, originally developed to reconstruct training data of DNN image classifiers. We make the key observation that a DLG attack, applied to our setting, infers the average location of a batch of local data, and can thus be used to reconstruct the target users' trajectory at a coarse granularity. We build on this observation to protect location privacy, in our setting, by revisiting and designing mechanisms within the federated learning framework including: tuning the FL parameters for averaging, curating local batches so as to mislead the DLG attacker, and aggregating across multiple users with different trajectories. We evaluate the performance of our algorithms through both analysis and simulation based on real-world mobile datasets, and we show that they achieve a good privacy-utility tradeoff.

cs.LG↗

arXiv2023

Efficient Toxic Content Detection by Bootstrapping and Distilling Large Language Models

Toxic content detection is crucial for online services to remove inappropriate content that violates community standards. To automate the detection process, prior works have proposed varieties of machine learning (ML) approaches to train Language Models (LMs) for toxic content detection. However, both their accuracy and transferability across datasets are limited. Recently, Large Language Models (LLMs) have shown promise in toxic content detection due to their superior zero-shot and few-shot in-context learning ability as well as broad transferability on ML tasks. However, efficiently designing prompts for LLMs remains challenging. Moreover, the high run-time cost of LLMs may hinder their deployments in production. To address these challenges, in this work, we propose BD-LLM, a novel and efficient approach to Bootstrapping and Distilling LLMs for toxic content detection. Specifically, we design a novel prompting method named Decision-Tree-of-Thought (DToT) to bootstrap LLMs' detection performance and extract high-quality rationales. DToT can automatically select more fine-grained context to re-prompt LLMs when their responses lack confidence. Additionally, we use the rationales extracted via DToT to fine-tune student LMs. Our experimental results on various datasets demonstrate that DToT can improve the accuracy of LLMs by up to 4.6%. Furthermore, student LMs fine-tuned with rationales extracted via DToT outperform baselines on all datasets with up to 16.9\% accuracy improvement, while being more than 60x smaller than conventional LLMs. Finally, we observe that student LMs fine-tuned with rationales exhibit better cross-dataset transferability.

cs.CL↗

arXiv2023

Finding emergence in data by maximizing effective information

Quantifying emergence and modeling emergent dynamics in a data-driven manner for complex dynamical systems is challenging due to the lack of direct observations at the micro-level. Thus, it's crucial to develop a framework to identify emergent phenomena and capture emergent dynamics at the macro-level using available data. Inspired by the theory of causal emergence (CE), this paper introduces a machine learning framework to learn macro-dynamics in an emergent latent space and quantify the degree of CE. The framework maximizes effective information, resulting in a macro-dynamics model with enhanced causal effects. Experimental results on simulated and real data demonstrate the effectiveness of the proposed framework. It quantifies degrees of CE effectively under various conditions and reveals distinct influences of different noise types. It can learn a one-dimensional coarse-grained macro-state from fMRI data, to represent complex neural activities during movie clip viewing. Furthermore, improved generalization to different test environments is observed across all simulation data.

physics.soc-ph↗