Search

Understanding and controlling non-equilibrium processes at ultralow temperatures are central to quantum physics and technology. In such extreme environments, quantum coherence and dissipation can interact intimately to give rise to intriguing thermalization phenomena. Here, we experimentally and theoretically demonstrate a novel scenario of thermalization in an ultracold atomic system, distinct from various quantum thermalization scenarios currently under intense investigations. We observe that after a sudden quench, an atomic Bose-Einstein condensate (BEC) behaves as a rigid body and undergoes a random walk in momentum space due to atom loss and interactions with the surrounding thermal component. Consequently its center of mass degree of freedom gets heated up at a constant rate. This heating mechanism, rooted in random momentum scattering, falls into the paradigm of stochastic heating initiated by Fermi and thoroughly explored in plasma physics, which differs conceptually from the traditional thermal conduction. At longer times, the stochastic heating of the BEC is balanced by forced evaporative cooling, and a Maxwell-Boltzmann distribution is achieved. Our findings offer new perspectives on the non-equilibrium dynamics of open Bose systems at ultralow temperature and quantum thermalization., Comment: 9 pages, 6 figures

SO(2,1) dynamical symmetry makes a remarkable prediction that the breathing oscillation of a scale invariant quantum gas in an isotropic harmonic trap is isentropic and can persist indefinitely. In 2D, this symmetry is broken due to quantum anomaly in the strongly interacting range, and consequently the lifetime of the breathing mode becomes finite. The persistent breather in a strongly interacting system has so far not been realized. Here we experimentally achieve the long-lived breathing mode in a 3D unitary Fermi gas, which is protected by the SO(2,1) symmetry. The nearly perfect SO(2,1) symmetry is realized by loading the ultracold Fermi gas in an isotropic trap and tuning the interatomic interaction to resonance. The breathing mode oscillates at twice the trapping frequency even for large excitation amplitudes. The ratio of damping rate to oscillation frequency is as small as 0.002, providing an interacting persistent breather. The oscillation frequency and damping rate keep nearly constant for different atomic densities and temperatures, demonstrating the robustness of the SO(2,1) symmetry in 3D. The factors that lead to the residual damping have also been clarified. This work opens the way to study many-body non-equilibrium dynamics related to the dynamical symmetry., Comment: 12 pages, 7 figures

Nonlinear spin dynamics are essential in exploring nonequilibrium quantum phenomena and have broad applications in precision measurement. Among these systems, the combination of a bias magnetic field and feedback mechanisms can induce self-sustained oscillations at the base Larmor frequency due to nonlinearity. These features have driven the development of single-species and multiple-species spin masers. The latter, with multiple discrete Larmor frequencies, provides significant advantages for precision measurement by mitigating uncertainties in precession frequencies due to long-term drifts in experimental conditions. The self-sustained oscillations of single-species and multiple-species spin masers correspond to limit cycles and quasi-periodic orbits of the stable nonlinear dynamics of the systems respectively; the correspondence is elucidated in a recent study on a related spin system featuring two discrete intrinsic Larmor frequencies under dual bias magnetic fields. Here, we extend the study to the case that the intrinsic Larmor frequencies of individual spins of the system, given rise to by an inhomogeneous bias magnetic field, form a continuum. We show that generically the stable dynamics of the system includes limit cycles, quasi-periodic orbits, and chaos. We establish the relation between the synchronization frequency of limit cycles and the field inhomogeneity and derive an equation determining the stability of limit cycles. Furthermore, detailed characteristics of different dynamical phases, especially the robustness of limit cycles and quasi-periodic orbits against experimental fluctuations, are discussed. Our findings not only encompass the case of discrete Larmor frequencies, but also provide crucial insights for precision measurement and the exploration of continuous time crystals and quasi-crystals.

A new method is presented to generate atomic structures that reproduce the essential characteristics of arbitrary material systems, phases, or ensembles. Previous methods allow one to reproduce the essential characteristics (e.g. chemical disorder) of a large random alloy within a small crystal structure. The ability to generate small representations of random alloys, with the restriction to crystal systems, results from using the fixed-lattice cluster correlations to describe structural characteristics. A more general description of the structural characteristics of atomic systems is obtained using complete sets of atomic environment descriptors. These are used within for generating representative atomic structures without restriction to fixed lattices. A general data-driven approach is provided utilizing the atomic cluster expansion(ACE) basis. The N-body ACE descriptors are a complete set of atomic environment descriptors that span both chemical and spatial degrees of freedom and are used within for describing atomic structures. The generalized representative structure(GRS) method presented within generates small atomic structures that reproduce ACE descriptor distributions corresponding to arbitrary structural and chemical complexity. It is shown that systematically improvable representations of crystalline systems on fixed parent lattices, amorphous materials, liquids, and ensembles of atomic structures may be produced efficiently through optimization algorithms. We highlight reduced representations of atomistic machine-learning training datasets that contain similar amounts of information and small 40-72 atom representations of liquid phases. The ability to use GRS methodology as a driver for informed novel structure generation is also demonstrated. The advantages over other data-driven methods and state-of-the-art methods restricted to high-symmetry systems are highlighted.

Combining multiple predictors obtained from distributed data sources to an accurate meta-learner is promising to achieve enhanced performance in lots of prediction problems. As the accuracy of each predictor is usually unknown, integrating the predictors to achieve better performance is challenging. Conventional ensemble learning methods assess the accuracy of predictors based on extensive labeled data. In practical applications, however, the acquisition of such labeled data can prove to be an arduous task. Furthermore, the predictors under consideration may exhibit high degrees of correlation, particularly when similar data sources or machine learning algorithms were employed during their model training. In response to these challenges, this paper introduces a novel structured unsupervised ensemble learning model (SUEL) to exploit the dependency between a set of predictors with continuous predictive scores, rank the predictors without labeled data and combine them to an ensembled score with weights. Two novel correlation-based decomposition algorithms are further proposed to estimate the SUEL model, constrained quadratic optimization (SUEL.CQO) and matrix-factorization-based (SUEL.MF) approaches. The efficacy of the proposed methods is rigorously assessed through both simulation studies and real-world application of risk genes discovery. The results compellingly demonstrate that the proposed methods can efficiently integrate the dependent predictors to an ensemble model without the need of ground truth data.

We investigate the scattering properties and bound states of a quasi-two-dimensional (q2D) spin-polarized Fermi gas near a $p$-wave Feshbach resonance. Strong confinement promotes the out-of-plane spatial wave functions to a discrete, gapped orbital degree of freedom. Exchange-antisymmetric orbital pair wave functions are predicted to give rise to low-energy q2D interactions with $s$-wave symmetry. Using radiofrequency (rf) spectroscopy, we observe the signature power-law scaling and the dimensional-crossover feature anticipated for the emergent $s$-wave channel. Additionally, we demonstrate that two types of low-energy dimers, with either $s$-wave and $p$-wave symmetry, could be formed via rf spin-flip association from an orbital mixture. These findings illustrate how gapped orbital degrees of freedom can provide additional control over scattering symmetries in strongly confined ultracold gases.

Low-dimensional ultracold gases are created in the laboratory by confining three-dimensional (3D) gases inside highly anisotropic trapping potentials. Such trap geometries not only provide access to simulating one-dimensional (1D) and two-dimensional (2D) physics, but also can be used to study how the system crosses over towards a 3D system in the limit of weak confinement. In this work, we study the signature in radio-frequency (RF) spectroscopy for both the 1D-to-3D and the 2D-to-3D crossovers, in spin-polarized Fermi gases. We solve the two-body scattering T-matrix in the presence of strong harmonic confinement and use it to evaluate the two-body bound state and the RF spectroscopy transfer rate in the high frequency limit, covering both the quasi-low-dimensional and 3D limits. We find that in order to understand the dimensional crossover for spin-polarized Fermi gases with p-wave interactions, one needs to take into account an emergent s-wave interaction., Comment: 16 pages, 5 figures, 4 appendices

We present Polynomial Attention Drop-in Replacement (PADRe), a novel and unifying framework designed to replace the conventional self-attention mechanism in transformer models. Notably, several recent alternative attention mechanisms, including Hyena, Mamba, SimA, Conv2Former, and Castling-ViT, can be viewed as specific instances of our PADRe framework. PADRe leverages polynomial functions and draws upon established results from approximation theory, enhancing computational efficiency without compromising accuracy. PADRe's key components include multiplicative nonlinearities, which we implement using straightforward, hardware-friendly operations such as Hadamard products, incurring only linear computational and memory costs. PADRe further avoids the need for using complex functions such as Softmax, yet it maintains comparable or superior accuracy compared to traditional self-attention. We assess the effectiveness of PADRe as a drop-in replacement for self-attention across diverse computer vision tasks. These tasks include image classification, image-based 2D object detection, and 3D point cloud object detection. Empirical results demonstrate that PADRe runs significantly faster than the conventional self-attention (11x ~ 43x faster on server GPU and mobile NPU) while maintaining similar accuracy when substituting self-attention in the transformer models.

Online kernel selection is a fundamental problem of online kernel methods.In this paper,we study online kernel selection with memory constraint in which the memory of kernel selection and online prediction procedures is limited to a fixed budget. An essential question is what is the intrinsic relationship among online learnability, memory constraint, and data complexity? To answer the question,it is necessary to show the trade-offs between regret and memory constraint.Previous work gives a worst-case lower bound depending on the data size,and shows learning is impossible within a small memory constraint.In contrast, we present distinct results by offering data-dependent upper bounds that rely on two data complexities:kernel alignment and the cumulative losses of competitive hypothesis.We propose an algorithmic framework giving data-dependent upper bounds for two types of loss functions.For the hinge loss function,our algorithm achieves an expected upper bound depending on kernel alignment.For smooth loss functions,our algorithm achieves a high-probability upper bound depending on the cumulative losses of competitive hypothesis.We also prove a matching lower bound for smooth loss functions.Our results show that if the two data complexities are sub-linear,then learning is possible within a small memory constraint.Our algorithmic framework depends on a new buffer maintaining framework and a reduction from online kernel selection to prediction with expert advice. Finally,we empirically verify the prediction performance of our algorithms on benchmark datasets.

In this paper, we develop the appropriate set of hydrodynamic equations in a U(N) invariant superfluid that couple the dynamics of superflow and magnetization. In the special case when both the superfluid and normal velocities are zero, the hydrodynamic equations reduce to a generalized version of Landau-Lifshitz equation for ferromagnetism with U(N) symmetry. When both velocities are non-zero, there appears couplings between the superflow and magnetization dynamics, and the superfluid velocity no longer satisfies the irrotational condition. On the other hand, the magnitude of magnetization is no longer a constant of motion as was the case for the standard Landau-Lifshitz theory. In comparison with the simple superfluid, the first sound and second sounds are modified by a non-zero magnetization through various thermodynamic functions. For U(2) invariant superfluid, we get both (zero-) sound wave and a spin wave at zero temperature. It is found that the dispersion of spin wave is always quadratic, which is consistent with detailed microscopic analysis. In the Appendix, we show that the hydrodynamic equation for a U(N) invariant superfluid can be obtained from the general hydrodynamic equation with arbitrary internal symmetries.

In this paper, we propose a novel video depth estimation approach, FutureDepth, which enables the model to implicitly leverage multi-frame and motion cues to improve depth estimation by making it learn to predict the future at training. More specifically, we propose a future prediction network, F-Net, which takes the features of multiple consecutive frames and is trained to predict multi-frame features one time step ahead iteratively. In this way, F-Net learns the underlying motion and correspondence information, and we incorporate its features into the depth decoding process. Additionally, to enrich the learning of multiframe correspondence cues, we further leverage a reconstruction network, R-Net, which is trained via adaptively masked auto-encoding of multiframe feature volumes. At inference time, both F-Net and R-Net are used to produce queries to work with the depth decoder, as well as a final refinement network. Through extensive experiments on several benchmarks, i.e., NYUDv2, KITTI, DDAD, and Sintel, which cover indoor, driving, and open-domain scenarios, we show that FutureDepth significantly improves upon baseline models, outperforms existing video depth estimation methods, and sets new state-of-the-art (SOTA) accuracy. Furthermore, FutureDepth is more efficient than existing SOTA video depth estimation models and has similar latencies when comparing to monocular models

Empathetic response generation is increasingly significant in AI, necessitating nuanced emotional and cognitive understanding coupled with articulate response expression. Current large language models (LLMs) excel in response expression; however, they lack the ability to deeply understand emotional and cognitive nuances, particularly in pinpointing fine-grained emotions and their triggers. Conversely, small-scale empathetic models (SEMs) offer strength in fine-grained emotion detection and detailed emotion cause identification. To harness the complementary strengths of both LLMs and SEMs, we introduce a Hybrid Empathetic Framework (HEF). HEF regards SEMs as flexible plugins to improve LLM's nuanced emotional and cognitive understanding. Regarding emotional understanding, HEF implements a two-stage emotion prediction strategy, encouraging LLMs to prioritize primary emotions emphasized by SEMs, followed by other categories, substantially alleviates the difficulties for LLMs in fine-grained emotion detection. Regarding cognitive understanding, HEF employs an emotion cause perception strategy, prompting LLMs to focus on crucial emotion-eliciting words identified by SEMs, thus boosting LLMs' capabilities in identifying emotion causes. This collaborative approach enables LLMs to discern emotions more precisely and formulate empathetic responses. We validate HEF on the Empathetic-Dialogue dataset, and the findings indicate that our framework enhances the refined understanding of LLMs and their ability to convey empathetic responses., Comment: 12 pages, 4 figures

Self-supervised methods have gained prominence in time series anomaly detection due to the scarcity of available annotations. Nevertheless, they typically demand extensive training data to acquire a generalizable representation map, which conflicts with scenarios of a few available samples, thereby limiting their performance. To overcome the limitation, we propose \textbf{AnomalyLLM}, a knowledge distillation-based time series anomaly detection approach where the student network is trained to mimic the features of the large language model (LLM)-based teacher network that is pretrained on large-scale datasets. During the testing phase, anomalies are detected when the discrepancy between the features of the teacher and student networks is large. To circumvent the student network from learning the teacher network's feature of anomalous samples, we devise two key strategies. 1) Prototypical signals are incorporated into the student network to consolidate the normal feature extraction. 2) We use synthetic anomalies to enlarge the representation gap between the two networks. AnomalyLLM demonstrates state-of-the-art performance on 15 datasets, improving accuracy by at least 14.5\% in the UCR dataset., Comment: 12 pages, 5 figures

In this Letter, we discuss the effects of $p$-wave attractive interaction in a spin-$1/2$ Bose gas. With a repulsive $s$-wave background interaction, we show that for weak $p$-wave attraction, one obtains a standard Bose-Einstein condensate at zero momentum with spins fully polarized. Upon increasing the $p$-wave attraction, a fragmented condensate state with singlet pair formation and $p$-wave correlation emerges. We determine the transition point and investigate the properties of the fragmented condensate using an ansatz wave function. We construct the relevant Gross-Pitaevskii equations for the fragmented condensate and show that the sound velocities are anisotropic and may vanish in specific directions. Based on the many-body wave function, we also discuss the low-energy spin excitations of the system.

Time series anomaly detection (TSAD) plays a vital role in various domains such as healthcare, networks, and industry. Considering labels are crucial for detection but difficult to obtain, we turn to TSAD with inexact supervision: only series-level labels are provided during the training phase, while point-level anomalies are predicted during the testing phase. Previous works follow a traditional multi-instance learning (MIL) approach, which focuses on encouraging high anomaly scores at individual time steps. However, time series anomalies are not only limited to individual point anomalies, they can also be collective anomalies, typically exhibiting abnormal patterns over subsequences. To address the challenge of collective anomalies, in this paper, we propose a tree-based MIL framework (TreeMIL). We first adopt an N-ary tree structure to divide the entire series into multiple nodes, where nodes at different levels represent subsequences with different lengths. Then, the subsequence features are extracted to determine the presence of collective anomalies. Finally, we calculate point-level anomaly scores by aggregating features from nodes at different levels. Experiments conducted on seven public datasets and eight baselines demonstrate that TreeMIL achieves an average 32.3% improvement in F1- score compared to previous state-of-the-art methods. The code is available at https://github.com/fly-orange/TreeMIL., Comment: This paper has been accepted by IEEE ICASSP 2024

Leveraging multimodal data is an inherent requirement for comprehending geographic objects. However, due to the high heterogeneity in structure and semantics among various spatio-temporal modalities, the joint interpretation of multimodal spatio-temporal data has long been an extremely challenging problem. The primary challenge resides in striking a trade-off between the cohesion and autonomy of diverse modalities. This trade-off becomes progressively nonlinear as the number of modalities expands. Inspired by the human cognitive system and linguistic philosophy, where perceptual signals from the five senses converge into language, we introduce the Language as Reference Framework (LaRF), a fundamental principle for constructing a multimodal unified model. Building upon this, we propose AllSpark, a multimodal spatio-temporal general artificial intelligence model. Our model integrates ten different modalities into a unified framework. To achieve modal cohesion, AllSpark introduces a modal bridge and multimodal large language model (LLM) to map diverse modal features into the language feature space. To maintain modality autonomy, AllSpark uses modality-specific encoders to extract the tokens of various spatio-temporal modalities. Finally, observing a gap between the model's interpretability and downstream tasks, we designed modality-specific prompts and task heads, enhancing the model's generalization capability across specific tasks. Experiments indicate that the incorporation of language enables AllSpark to excel in few-shot classification tasks for RGB and point cloud modalities without additional training, surpassing baseline performance by up to 41.82\%. The source code is available at https://github.com/GeoX-Lab/AllSpark., Comment: 19 pages, 19 tables, 3 figures

In this paper, we study the mistake bound of online kernel learning on a budget. We propose a new budgeted online kernel learning model, called Ahpatron, which significantly improves the mistake bound of previous work and resolves the open problem posed by Dekel, Shalev-Shwartz, and Singer (2005). We first present an aggressive variant of Perceptron, named AVP, a model without budget, which uses an active updating rule. Then we design a new budget maintenance mechanism, which removes a half of examples,and projects the removed examples onto a hypothesis space spanned by the remaining examples. Ahpatron adopts the above mechanism to approximate AVP. Theoretical analyses prove that Ahpatron has tighter mistake bounds, and experimental results show that Ahpatron outperforms the state-of-the-art algorithms on the same or a smaller budget.

Wireless blockchain networks have played an important role in many network scenarios, among which wireless Practical Byzantine Fault Tolerance (PBFT) consensus is regarded as one of the most important consensus mechanisms. It enables nodes in wireless networks to reach consistency without any trusted entity. However, due to the instability of wireless communication links, the reliability of the PBFT consensus will be seriously affected. Meanwhile, it is difficult for nodes in wireless scenarios to obtain a timely energy supply. The high-energy-consumption blockchain functions will quickly consume the power of nodes, thus, affecting consensus performance. Fortunately, the symbiotic radio (SR) system enabled by cognitive backscatter communications can provide a solution to the above problems. In SR, the secondary transmitter (STx) transmits messages by modulating its information over the radio frequency (RF) signal of the primary transmitter (PTx) with extremely low energy consumption, and the STx can provide multipath gain to the PTx in return. In our paper, we propose the symbiotic PBFT (S-PBFT) consensus benefited from the mutualistic transmission in SR, which can increase the consensus security by 54.82%, and save energy consumption by about 10%., Comment: IEEE Global Communications Conference 2023