Search

Modern recommender systems aim to deeply understand users' complex preferences through their past interactions. While deep collaborative filtering approaches using Graph Neural Networks (GNNs) excel at capturing user-item relationships, their effectiveness is limited when handling sparse data or zero-shot scenarios, primarily due to constraints in ID-based embedding functions. To address these challenges, we propose a model-agnostic recommendation instruction-tuning paradigm that seamlessly integrates large language models with collaborative filtering. Our proposed $\underline{Rec}$ommendation $\underline{L}$anguage $\underline{M}$odel (RecLM) enhances the capture of user preference diversity through a carefully designed reinforcement learning reward function that facilitates self-augmentation of language models. Comprehensive evaluations demonstrate significant advantages of our approach across various settings, and its plug-and-play compatibility with state-of-the-art recommender systems results in notable performance enhancements. The implementation of our RecLM framework is publicly available at: https://github.com/HKUDS/RecLM. more...

Despite extensive microscale studies, the macroscopic mechanical properties of monolayer graphene remain underexplored. Here, we report the Young's modulus ($E$ = 1.11 $\pm$ 0.04 TPa), tensile strength (${\sigma}$ = 27.40 $\pm$ 4.36 GPa), and failure strain (${\epsilon}_f$ = 6.01 $\pm$ 0.92 %) of centimeter-scale single-crystal monolayer graphene (SCG) 'dog bone' samples with edges aligned along the zigzag (zz) direction, supported by an ultra-thin polymer (polycarbonate) film. For samples with edges along the armchair (ac) direction, we obtain $E$ = 1.01 $\pm$ 0.10 TPa, ${\sigma}$ = 20.21 $\pm$ 3.22 GPa, ${\epsilon}_f$ = 3.69 $\pm$ 0.38 %, and for chiral samples whose edges were between zz and ac, we obtain $E$= 0.75 $\pm$ 0.12 TPa, ${\sigma}$ = 23.56 $\pm$ 3.42 GPa, and ${\epsilon}_f$ = 4.53 $\pm$ 0.40 %. The SCG is grown on single crystal Cu(111) foils by chemical vapor deposition (CVD). We used a home-built 'float-on-water' (FOW) tensile testing system for tensile loading measurements that also enabled in situ crack observation. The quantized fracture mechanics (QFM) analysis predicts an edge defect size from several to tens of nanometers based on chirality and notch angle. Through Weibull analysis and given that the fatal defects are confined on the edges of macroscale samples, we projected strength ranging from 13.67 to 18.43 GPa for an A4-size SCG according to their chirality. The exceptional mechanical performance of macroscale single crystal graphene (SCG) paves the way for its widespread use in a very wide variety of applications., Comment: 76 pages, 44 figures more...

The rise of online multi-modal sharing platforms like TikTok and YouTube has enabled personalized recommender systems to incorporate multiple modalities (such as visual, textual, and acoustic) into user representations. However, addressing the challenge of data sparsity in these systems remains a key issue. To address this limitation, recent research has introduced self-supervised learning techniques to enhance recommender systems. However, these methods often rely on simplistic random augmentation or intuitive cross-view information, which can introduce irrelevant noise and fail to accurately align the multi-modal context with user-item interaction modeling. To fill this research gap, we propose a novel multi-modal graph diffusion model for recommendation called DiffMM. Our framework integrates a modality-aware graph diffusion model with a cross-modal contrastive learning paradigm to improve modality-aware user representation learning. This integration facilitates better alignment between multi-modal feature information and collaborative relation modeling. Our approach leverages diffusion models' generative capabilities to automatically generate a user-item graph that is aware of different modalities, facilitating the incorporation of useful multi-modal knowledge in modeling user-item interactions. We conduct extensive experiments on three public datasets, consistently demonstrating the superiority of our DiffMM over various competitive baselines. For open-sourced model implementation details, you can access the source codes of our proposed framework at: https://github.com/HKUDS/DiffMM . more...

We investigated the optimal control of a continuous variable system, focusing on entanglement generation in an optomechanical system without utilizing Fock basis cutoffs. Using the Krotov algorithm to optimize the dynamics of the covariance matrix, we illustrated how to design a control objective function to manipulate the dynamics of the system to generate a desirable target state. We showed that entanglement between the macroscopic mechanical mirror and the quantum optical cavity can be reliably generated through imposing the control on the detuning of the external laser field. It has be shown that the control may be still achieved when imposing spectral constraints on the external field to restrict it to low-frequency components. In addition, we systematically studies the effects of quantum control on non-Markovian open system dynamics. We observed that memory effects can play a beneficial role in mitigating the detrimental impact of environmental noises. Specifically, the entanglement generated shows reduced decay in the presence of these memory effects., Comment: 12 pages, 8 figures more...

We consider the Jaynes-Cummings (JC) model embedded in a structured environment, where the atom inside an optical cavity will be affected by a hierarchical environment consisting of the cavity and its environment. We propose several effective strategies to control and suppress the decoherence effects to protect the quantum coherence of the JC atom. We study the non-perturbative control of the system dynamics by means of the leakage elimination operators. We also investigate a full quantum state reversal scheme by engineering the system and its coupling to the bath via the Petz recovery map. Our findings conclude that, with the Petz recovery map, the dynamics of the JC atom can be fully recovered regardless of Markov or non-Markovian noises. Finally, we show that our quantum control and recovery methods are effective at protecting different aspects of the system coherence., Comment: 8 pages, 3 figures more...

We propose a protocol to overcome the shot noise limit and reach the Heisenberg scaling limit for parameter estimation by using quantum optimal control and a time-reversal strategy. Exemplified through the phase estimation, which can play an important role in quantum navigation and measurement, we show that the uncertainty arising from a photon number measurement of the system can saturate the assisted Cream\'er-Rao bound, independent of the phase being estimated. In a realistic case with photon loss, we show that the optimal estimation may still be attainable by optimal control and a projective measurement on an ancilla two-level system coupled to photonic modes., Comment: 7 pages, 5 figures more...

Non-Markovian effects of an open system dynamics are typically characterized by non-monotonic information flows from the system to its environment or information backflows from the environment to the system. By using a two-level system (TLS) coupled to a dissipative single-mode cavity, we show that the geometric decoherence of the open quantum system of interest can serve as a reliable witness of non-Markovian dynamics. This geometric approach can also reveal the finer details about the dynamics such as the time points where the non-Markovian behaviors come into operation. Specifically, we show that the divergence of the geometric decoherence factor of the TLS can be a sufficient condition for the non-Markovian dynamics. Remarkably, it can even become a necessary and sufficient condition in certain cases., Comment: 6 pages, 3 figures more...

Few-shot Relation Extraction (FSRE) aims to extract relational facts from a sparse set of labeled corpora. Recent studies have shown promising results in FSRE by employing Pre-trained Language Models (PLMs) within the framework of supervised contrastive learning, which considers both instances and label facts. However, how to effectively harness massive instance-label pairs to encompass the learned representation with semantic richness in this learning paradigm is not fully explored. To address this gap, we introduce a novel synergistic anchored contrastive pre-training framework. This framework is motivated by the insight that the diverse viewpoints conveyed through instance-label pairs capture incomplete yet complementary intrinsic textual semantics. Specifically, our framework involves a symmetrical contrastive objective that encompasses both sentence-anchored and label-anchored contrastive losses. By combining these two losses, the model establishes a robust and uniform representation space. This space effectively captures the reciprocal alignment of feature distributions among instances and relational facts, simultaneously enhancing the maximization of mutual information across diverse perspectives within the same relation. Experimental results demonstrate that our framework achieves significant performance enhancements compared to baseline models in downstream FSRE tasks. Furthermore, our approach exhibits superior adaptability to handle the challenges of domain shift and zero-shot relation extraction. Our code is available online at https://github.com/AONE-NLP/FSRE-SaCon. more...

GNN-based recommenders have excelled in modeling intricate user-item interactions through multi-hop message passing. However, existing methods often overlook the dynamic nature of evolving user-item interactions, which impedes the adaption to changing user preferences and distribution shifts in newly arriving data. Thus, their scalability and performances in real-world dynamic environments are limited. In this study, we propose GraphPro, a framework that incorporates parameter-efficient and dynamic graph pre-training with prompt learning. This novel combination empowers GNNs to effectively capture both long-term user preferences and short-term behavior dynamics, enabling the delivery of accurate and timely recommendations. Our GraphPro framework addresses the challenge of evolving user preferences by seamlessly integrating a temporal prompt mechanism and a graph-structural prompt learning mechanism into the pre-trained GNN model. The temporal prompt mechanism encodes time information on user-item interaction, allowing the model to naturally capture temporal context, while the graph-structural prompt learning mechanism enables the transfer of pre-trained knowledge to adapt to behavior dynamics without the need for continuous incremental training. We further bring in a dynamic evaluation setting for recommendation to mimic real-world dynamic scenarios and bridge the offline-online gap to a better level. Our extensive experiments including a large-scale industrial deployment showcases the lightweight plug-in scalability of our GraphPro when integrated with various state-of-the-art recommenders, emphasizing the advantages of GraphPro in terms of effectiveness, robustness and efficiency. The implementation details and source code of our GraphPro are available in the repository at https://github.com/HKUDS/GraphPro, Comment: Accepted by WWW'2024, full paper more...

Aspect-based sentiment classification is a crucial problem in fine-grained sentiment analysis, which aims to predict the sentiment polarity of the given aspect according to its context. Previous works have made remarkable progress in leveraging attention mechanism to extract opinion words for different aspects. However, a persistent challenge is the effective management of semantic mismatches, which stem from attention mechanisms that fall short in adequately aligning opinions words with their corresponding aspect in multi-aspect sentences. To address this issue, we propose a novel Aspect-oriented Opinion Alignment Network (AOAN) to capture the contextual association between opinion words and the corresponding aspect. Specifically, we first introduce a neighboring span enhanced module which highlights various compositions of neighboring words and given aspects. In addition, we design a multi-perspective attention mechanism that align relevant opinion information with respect to the given aspect. Extensive experiments on three benchmark datasets demonstrate that our model achieves state-of-the-art results. The source code is available at https://github.com/AONE-NLP/ABSA-AOAN., Comment: 8 pages, 5 figure, ECAI 2023 more...

Current sequential recommender systems are proposed to tackle the dynamic user preference learning with various neural techniques, such as Transformer and Graph Neural Networks (GNNs). However, inference from the highly sparse user behavior data may hinder the representation ability of sequential pattern encoding. To address the label shortage issue, contrastive learning (CL) methods are proposed recently to perform data augmentation in two fashions: (i) randomly corrupting the sequence data (e.g. stochastic masking, reordering); (ii) aligning representations across pre-defined contrastive views. Although effective, we argue that current CL-based methods have limitations in addressing popularity bias and disentangling of user conformity and real interest. In this paper, we propose a new Debiased Contrastive learning paradigm for Recommendation (DCRec) that unifies sequential pattern encoding with global collaborative relation modeling through adaptive conformity-aware augmentation. This solution is designed to tackle the popularity bias issue in recommendation systems. Our debiased contrastive learning framework effectively captures both the patterns of item transitions within sequences and the dependencies between users across sequences. Our experiments on various real-world datasets have demonstrated that DCRec significantly outperforms state-of-the-art baselines, indicating its efficacy for recommendation. To facilitate reproducibility of our results, we make our implementation of DCRec publicly available at: https://github.com/HKUDS/DCRec., Comment: This paper is accepted by WWW'2023 more...

We study the remote entanglement generation between macroscopic microwave magnon modes in a coupled cavity system. The cavities are connected via an optical fiber, which necessitates the use of a frequency conversion inside the cavity. The converter may be implemented via a rare-earth doped crystal acting like an effective three-level system. The entanglement dynamics of the system is analytically studied, and an active optimal control method is also proposed where one may generate maximally entangled Bell states on demand with a given evolution time. The system dynamics and its control have also been studied in a generic non-Markovian open system framework, and the generated entanglement is found to be robust against environmental noises., Comment: 9 pages, 7 figures more...

Recent experiments show the existence of collective decoherence in quantum systems. We study the possibility of quantum computation in decoherence free subspace which is robust against such kind of decoherence processes. This passive protection protocol can be especially advantageous for continuous quantum computation such as quantum annealers. As an example we propose to use decoherence protected adiabatic quantum computation for the Grover search problem. Our proposal contains only two-body interactions, making it feasible in near-term quantum devices. more...

Click-Through Rate (CTR) prediction, which aims to estimate the probability that a user will click an item, is an essential component of online advertising. Existing methods mainly attempt to mine user interests from users' historical behaviours, which contain users' directly interacted items. Although these methods have made great progress, they are often limited by the recommender system's direct exposure and inactive interactions, and thus fail to mine all potential user interests. To tackle these problems, we propose Neighbor-Interaction based CTR prediction (NI-CTR), which considers this task under a Heterogeneous Information Network (HIN) setting. In short, Neighbor-Interaction based CTR prediction involves the local neighborhood of the target user-item pair in the HIN to predict their linkage. In order to guide the representation learning of the local neighbourhood, we further consider different kinds of interactions among the local neighborhood nodes from both explicit and implicit perspective, and propose a novel Graph-Masked Transformer (GMT) to effectively incorporates these kinds of interactions to produce highly representative embeddings for the target user-item pair. Moreover, in order to improve model robustness against neighbour sampling, we enforce a consistency regularization loss over the neighbourhood embedding. We conduct extensive experiments on two real-world datasets with millions of instances and the experimental results show that our proposed method outperforms state-of-the-art CTR models significantly. Meanwhile, the comprehensive ablation studies verify the effectiveness of every component of our model. Furthermore, we have deployed this framework on the WeChat Official Account Platform with billions of users. The online A/B tests demonstrate an average CTR improvement of 21.9 against all online baselines., Comment: 11 pages more...

We consider the quantum state control of a multi-state system which evolves an initial state into a target state. We explicitly demonstrate the control method in an interesting case involving the transfer and rotation of a Schr\"{o}dinger cat state through a coupled harmonic oscillator chain at a predetermined time $T$. We use the gradient-based Krotov's method to design the time-dependent parameters of the coupled chain to find an optimal control shape that will evolve the system into a target state. We show that the prescribed quantum state control can be achieved with high fidelity, and the robustness of the control against generic environment noises is explored. Our findings will be of interest for the optimal control of a many-body open quantum system in the presence of environmental noise., Comment: 12 pages, 6 figures more...

We propose and analyze a nonunitary variant of the continuous time Grover search algorithm based on frequent Zeno-type measurements. We show that the algorithm scales similarly to the pure quantum version by deriving tight analytical lower bounds on its efficiency for arbitrary database sizes and measurement parameters. We also study the behavior of the algorithm subject to noise, and find that under certain oracle and operational errors our measurement-based algorithm outperforms the standard algorithm, showing robustness against these noises. Our analysis is based on deriving a non-hermitian effective description of the algorithm, which yields a deeper insight into components responsible for the quantum and the classical operation of the protocol. more...

We investigate dynamical generation of macroscopic nonlocal entanglements between two remote massive magnon-superconducting-circuit hybrid systems. Two fiber-coupled microwave cavities are employed to serve as an interaction channel connecting two sets of macroscopic hybrid units each containing a magnon (hosted by a Yttrium-Iron-Garnet sphere) and a superconducting-circuit qubit. Surprisingly, it is found that stronger coupling does not necessarily mean faster entanglement generation. The proposed hybrid system allows the existence of an optimal fiber coupling strength that requests the shortest amount of time to generate a systematic maximal entanglement. Our theoretical results are shown to be within the scope of specific parameters that can be achieved with current technology. The noise effects on the implementation of systems are also treated in a general environment suggesting the robustness of entanglement generation. Our discrete-variable qubit-like entanglement theory of magnons may lead to direct applications in various quantum information tasks., Comment: 10 pages, 4 figures more...

We demonstrate terahertz time-domain spectroscopy (THz-TDS) to be an accurate, rapid and scalable method to probe the interaction-induced Fermi velocity renormalization {\nu}F^* of charge carriers in graphene. This allows the quantitative extraction of all electrical parameters (DC conductivity {\sigma}DC, carrier density n, and carrier mobility {\mu}) of large-scale graphene films placed on arbitrary substrates via THz-TDS. Particularly relevant are substrates with low relative permittivity (< 5) such as polymeric films, where notable renormalization effects are observed even at relatively large carrier densities (> 10^12 cm-2, Fermi level > 0.1 eV). From an application point of view, the ability to rapidly and non-destructively quantify and map the electrical ({\sigma}DC, n, {\mu}) and electronic ({\nu}F^* ) properties of large-scale graphene on generic substrates is key to utilize this material in applications such as metrology, flexible electronics as well as to monitor graphene transfers using polymers as handling layers., Comment: 23 pages, 8 figures more...

The evolution of a quantum system towards thermal equilibrium is usually studied by approximate methods, which have their limits of validity and should be checked against analytically solvable models. In this paper, we propose an analytically solvable model to investigate the energy transfer between a bosonic bath and a fermionic chain which are nonlinearly-coupled to each other. The bosonic bath consists of an infinite collection of non-interacting bosonic modes, while the fermionic chain is represented by a chain of interacting fermions with nearest-neighbor interactions. We compare behaviors of the temperature-dependent energy current $J_{T}$ and temperature-independent energy current $J_{TI}$ for different bath configurations. With respect to the bath spectrum, $J_{T}$ decays exponentially for Lorentz-Drude type bath, which is the same as the conventional approximations. On the other hand, the decay rate is $1/t^{3}$ for Ohmic type and $1/t$ for white noise, which doesn't have conventional counterparts. For the temperature-independent current $J_{TI}$, the decay rate is divergent for the Lorentz-Drude type bath, $1/t^{4}$ for the Ohmic bath, and $1/t$ for the white noise. When further considering the dynamics of the fermionic chain, the current will be modulated based on the envelope from the bath. As an example, for a bosonic bath with Ohmic spectrum, when the fermionic chain is uniformly-coupled, we have $J_{T}\propto1/t^{6}$ and $J_{TI}\propto1/t^{3}$. Remarkably, for perfect state transfer (PST) couplings, there always exists an oscillating quantum energy current $J_{TI}$. Moreover, it is interesting that $J_{T}$ is proportional to $(N-1)^{1/2}$ at certain times for PST couplings under Lorentz-Drude or Ohmic bath., Comment: 6 pages, 1 figures more...

Based on a generic quantum open system model, we study the geometric nature of decoherence by defining a complex-valued geometric phase through stochastic pure states describing non-unitary, non-cyclic and non-adiabatic evolutions. The ensemble average of the complex geometric phases for the pure stochastic states yields a conventional geometric phase together with an amplitude factor. We show that the decoherence process described by the decaying amplitude can be a geometric quantity independent of the system's dynamics. It is a remarkable fact that the geometric phase of a quantum system can serve as an ideal realisation of quantum gates due to its robustness against dynamical errors, however, in this paper we show that, for some open quantum systems, a desirable geometric phase may be accompanied by an unwanted robust geometric decoherence factor. Two exactly solvable models are studied to demonstrate that, while the decoherence is a purely dynamical effect for a dephasing two-level model, the decoherence in a dissipative two-level model can be a geometric process. Finally, we show that such a geometric decoherence effect may be eliminated by a non-perturbative control scheme., Comment: 8 pages, 3 figures more...