Search

Classification models are expected to perform equally well for different classes, yet in practice, there are often large gaps in their performance. This issue of class bias is widely studied in cases of datasets with sample imbalance, but is relatively overlooked in balanced datasets. In this work, we introduce the concept of spectral imbalance in features as a potential source for class disparities and study the connections between spectral imbalance and class bias in both theory and practice. To build the connection between spectral imbalance and class gap, we develop a theoretical framework for studying class disparities and derive exact expressions for the per-class error in a high-dimensional mixture model setting. We then study this phenomenon in 11 different state-of-the-art pretrained encoders and show how our proposed framework can be used to compare the quality of encoders, as well as evaluate and combine data augmentation strategies to mitigate the issue. Our work sheds light on the class-dependent effects of learning, and provides new insights into how state-of-the-art pretrained features may have unknown biases that can be diagnosed through their spectra., Comment: 25 pages, 9 figures

Deep Neural Nets (DNNs) learn latent representations induced by their downstream task, objective function, and other parameters. The quality of the learned representations impacts the DNN's generalization ability and the coherence of the emerging latent space. The Information Bottleneck (IB) provides a hypothetically optimal framework for data modeling, yet it is often intractable. Recent efforts combined DNNs with the IB by applying VAE-inspired variational methods to approximate bounds on mutual information, resulting in improved robustness to adversarial attacks. This work introduces a new and tighter variational bound for the IB, improving performance of previous IB-inspired DNNs. These advancements strengthen the case for the IB and its variational approximations as a data modeling framework, and provide a simple method to significantly enhance the adversarial robustness of classifier DNNs., Comment: 10 pages, 5 figures, code included in github repo

The recent discovery of possible high temperature superconductivity in single crystals of $La_3Ni_2O_7$ under pressure renews the interest in research on nickelates. The DFT calculations reveal that both $d_{z^2}$ and $d_{x^2-y^2}$ orbitals are active, which suggests a minimal two-orbital model to capture the low-energy physics of this system. In this work, we study a bilayer two-orbital $t-J$ model within multiband Gutzwiller approximation, and discuss the magnetism as well as the superconductivity over a wide range of the hole doping. Owing to the inter-orbital super-exchange process between $d_{z^2}$ and $d_{x^2-y^2}$ orbitals, the induced ferromagnetic coupling within layers competes with the conventional antiferromagnetic coupling, and leads to complicated hole doping dependence for the magnetic properties in the system. With increasing hole doping, the system transfers to A-AFM state from the starting G-AFM state. We also find the inter-layer superconducting pairing of $d_{x^2-y^2}$ orbitals dominates due to the large hopping parameter of $d_{z^2}$ along the vertical inter-layer bonds and significant Hund's coupling between $d_{z^2}$ and $d_{x^2-y^2}$ orbitals. Meanwhile, the G-AFM state and superconductivity state can coexist in the low hole doping regime. To take account of the pressure, we also analyze the impacts of inter-layer hopping amplitude on the system properties., Comment: 11 pages, 4 figures

We present GALA3D, generative 3D GAussians with LAyout-guided control, for effective compositional text-to-3D generation. We first utilize large language models (LLMs) to generate the initial layout and introduce a layout-guided 3D Gaussian representation for 3D content generation with adaptive geometric constraints. We then propose an instance-scene compositional optimization mechanism with conditioned diffusion to collaboratively generate realistic 3D scenes with consistent geometry, texture, scale, and accurate interactions among multiple objects while simultaneously adjusting the coarse layout priors extracted from the LLMs to align with the generated scene. Experiments show that GALA3D is a user-friendly, end-to-end framework for state-of-the-art scene-level 3D content generation and controllable editing while ensuring the high fidelity of object-level entities within the scene. The source codes and models will be available at gala3d.github.io.

Uranium ditelluride (UTe$_2$) is recognized as a host material to unconventional spin-triplet superconductivity, but it also exhibits a wealth of additional unusual behavior at high magnetic fields. One of the most prominent signatures of the unconventional superconductivity is a large and anisotropic upper critical field that exceeds the paramagnetic limit. This superconductivity survives to 35 T and is bounded by a discontinuous magnetic transition, which itself is also field-direction-dependent. A different, reentrant superconducting phase emerges only on the high-field side of the magnetic transition, in a range of angles between the crystallographic $b$ and $c$ axes. This review discusses the current state of knowledge of these high-field phases, the high-field behavior of the heavy fermion normal state, and other phases that are stabilized by applied pressure., Comment: 34 pages, 12 figures

Fast and high-fidelity qubit initialization is crucial for low-frequency qubits such as fluxonium, and in applications of many quantum algorithms and quantum error correction codes. In a circuit quantum electrodynamics system, the initialization is typically achieved by transferring the state between the qubit and a short-lived cavity through microwave driving, also known as the sideband cooling process in atomic system. Constrained by the selection rules from the parity symmetry of the wavefunctions, the sideband transitions are only enabled by multi-photon processes which requires multi-tone or strong driving. Leveraging the flux-tunability of fluxonium, we circumvent this limitation by breaking flux symmetry to enable an interaction between a non-computational qubit transition and the cavity excitation. With single-tone sideband driving, we realize qubit initialization with a fidelity exceeding 99% within a duration of 300 ns, robust against the variation of control parameters. Furthermore, we show that our initialization scheme has a built-in benefit in simultaneously removing the second-excited state population of the qubit, and can be easily incorporated into a large-scale fluxonium processor.

Soft robots have immense potential given their inherent safety and adaptability, but challenges in soft actuator forces and design constraints have limited scaling up soft robots to larger sizes. Electrothermal shape memory alloy (SMA) artificial muscles have the potential to create these large forces and high displacements, but consistently using these muscles under a well-defined model, in-situ in a soft robot, remains an open challenge. This article provides a system for maintaining the highest-possible consistent SMA forces, over long lifetimes, by combining a fatigue testing protocol with a supervisory control system for the muscles' internal temperature state. We propose a design of a soft limb with swap-able SMA muscles, and deploy the limb in a blocked-force test to quantify the relationship between the measured maximum force at different temperatures over different lifetimes. Then, by applying an invariance-based control system to maintain temperatures under our long-life limit, we demonstrate consistent high forces in a practical task over hundreds of cycles. The method we developed allows for practical implementation of SMAs in soft robots through characterizing and controlling their behavior in-situ, and provides a method to impose limits that maximize their consistent, repeatable behavior., Comment: 8 pages, 8 figures, accepted by 2024 IEEE International Conference on Soft Robotics (RoboSoft)

The development of artificial intelligence systems is transitioning from creating static, task-specific models to dynamic, agent-based systems capable of performing well in a wide range of applications. We propose an Interactive Agent Foundation Model that uses a novel multi-task agent training paradigm for training AI agents across a wide range of domains, datasets, and tasks. Our training paradigm unifies diverse pre-training strategies, including visual masked auto-encoders, language modeling, and next-action prediction, enabling a versatile and adaptable AI framework. We demonstrate the performance of our framework across three separate domains -- Robotics, Gaming AI, and Healthcare. Our model demonstrates its ability to generate meaningful and contextually relevant outputs in each area. The strength of our approach lies in its generality, leveraging a variety of data sources such as robotics sequences, gameplay data, large-scale video datasets, and textual information for effective multimodal and multi-task learning. Our approach provides a promising avenue for developing generalist, action-taking, multimodal systems.

Natural language serves as a common and straightforward control signal for humans to interact seamlessly with machines. Recognizing the importance of this interface, the machine learning community is investing considerable effort in generating data that is semantically coherent with textual instructions. While strides have been made in text-to-data generation spanning image editing, audio synthesis, video creation, and beyond, low-resource areas characterized by expensive annotations or complex data structures, such as molecules, motion dynamics, and time series, often lack textual labels. This deficiency impedes supervised learning, thereby constraining the application of advanced generative models for text-to-data tasks. In response to these challenges in the low-resource scenario, we propose Text2Data, a novel approach that utilizes unlabeled data to understand the underlying data distribution through an unsupervised diffusion model. Subsequently, it undergoes controllable finetuning via a novel constraint optimization-based learning objective that ensures controllability and effectively counteracts catastrophic forgetting. Comprehensive experiments demonstrate that Text2Data is able to achieve enhanced performance regarding controllability across various modalities, including molecules, motions and time series, when compared to existing baselines., Comment: We propose a method that can achieve text-to-data generation under low-resource situation

Visually Rich Form Understanding (VRFU) poses a complex research problem due to the documents' highly structured nature and yet highly variable style and content. Current annotation schemes decompose form understanding and omit key hierarchical structure, making development and evaluation of end-to-end models difficult. In this paper, we propose a novel F1 metric to evaluate form parsers and describe a new content-agnostic, tree-based annotation scheme for VRFU: TreeForm. We provide methods to convert previous annotation schemes into TreeForm structures and evaluate TreeForm predictions using a modified version of the normalized tree-edit distance. We present initial baselines for our end-to-end performance metric and the TreeForm edit distance, averaged over the FUNSD and XFUND datasets, of 61.5 and 26.4 respectively. We hope that TreeForm encourages deeper research in annotating, modeling, and evaluating the complexities of form-like documents.

Microcalorimeters that leverage microwave kinetic inductance detectors to read out phonon signals in the particle-absorbing target, referred to as kinetic inductance phonon-mediated (KIPM) detectors, offer an attractive detector architecture to probe dark matter (DM) down to the fermionic thermal relic mass limit. A prototype KIPM detector featuring a single aluminum resonator patterned onto a 1-gram silicon substrate was operated in the NEXUS low-background facility at Fermilab for characterization and evaluation of this detector architecture's efficacy for a dark matter search. An energy calibration was performed by exposing the bare substrate to a pulsed source of 470 nm photons, resulting in a baseline resolution on the energy absorbed by the phonon sensor of $2.1\pm0.2$ eV, a factor of two better than the current state-of-the-art, enabled by millisecond-scale quasiparticle lifetimes. However, due to the sub-percent phonon collection efficiency, the resolution on energy deposited in the substrate is limited to $\sigma_E=318 \pm 28$ eV. We further model the signal pulse shape as a function of device temperature to extract quasiparticle lifetimes, as well as the observed noise spectra, both of which impact the baseline resolution of the sensor.

Contrastive Language-Image Pretraining (CLIP) has gained popularity for its remarkable zero-shot capacity. Recent research has focused on developing efficient fine-tuning methods, such as prompt learning and adapter, to enhance CLIP's performance in downstream tasks. However, these methods still require additional training time and computational resources, which is undesirable for devices with limited resources. In this paper, we revisit a classical algorithm, Gaussian Discriminant Analysis (GDA), and apply it to the downstream classification of CLIP. Typically, GDA assumes that features of each class follow Gaussian distributions with identical covariance. By leveraging Bayes' formula, the classifier can be expressed in terms of the class means and covariance, which can be estimated from the data without the need for training. To integrate knowledge from both visual and textual modalities, we ensemble it with the original zero-shot classifier within CLIP. Extensive results on 17 datasets validate that our method surpasses or achieves comparable results with state-of-the-art methods on few-shot classification, imbalanced learning, and out-of-distribution generalization. In addition, we extend our method to base-to-new generalization and unsupervised learning, once again demonstrating its superiority over competing approaches. Our code is publicly available at \url{https://github.com/mrflogs/ICLR24}., Comment: Accepted by ICLR 2024

With the emergence of pretrained vision-language models (VLMs), considerable efforts have been devoted to fine-tuning them for downstream tasks. Despite the progress made in designing efficient fine-tuning methods, such methods require access to the model's parameters, which can be challenging as model owners often opt to provide their models as a black box to safeguard model ownership. This paper proposes a \textbf{C}ollabo\textbf{ra}tive \textbf{F}ine-\textbf{T}uning (\textbf{CraFT}) approach for fine-tuning black-box VLMs to downstream tasks, where one only has access to the input prompts and the output predictions of the model. CraFT comprises two modules, a prompt generation module for learning text prompts and a prediction refinement module for enhancing output predictions in residual style. Additionally, we introduce an auxiliary prediction-consistent loss to promote consistent optimization across these modules. These modules are optimized by a novel collaborative training algorithm. Extensive experiments on few-shot classification over 15 datasets demonstrate the superiority of CraFT. The results show that CraFT achieves a decent gain of about 12\% with 16-shot datasets and only 8,000 queries. Moreover, CraFT trains faster and uses only about 1/80 of the memory footprint for deployment, while sacrificing only 1.62\% compared to the white-box method. Our code is publicly available at https://github.com/mrflogs/CraFT ., Comment: Accepted by ICML 2024

This paper considers a class of noncooperative games in which the feasible decision sets of all players are coupled together by a coupled inequality constraint. Adopting the variational inequality formulation of the game, we first introduce a new local edge-based equilibrium condition and develop a distributed primal-dual proximal algorithm with full information. Considering challenges when communication delays occur, we devise an asynchronous distributed algorithm to seek a generalized Nash equilibrium. This asynchronous scheme arbitrarily activates one player to start new computations independently at different iteration instants, which means that the picked player can use the involved out-dated information from itself and its neighbors to perform new updates. A distinctive attribute is that the proposed algorithms enable the derivation of new distributed forward-backward-like extensions. In theoretical aspect, we provide explicit conditions on algorithm parameters, for instance, the step-sizes to establish a sublinear convergence rate for the proposed synchronous algorithm. Moreover, the asynchronous algorithm guarantees almost sure convergence in expectation under the same step-size conditions and some standard assumptions. An interesting observation is that our analysis approach improves the convergence rate of prior synchronous distributed forward-backward-based algorithms. Finally, the viability and performance of the proposed algorithms are demonstrated by numerical studies on the networked Cournot competition., Comment: 13 pages, 2 figures

In today's digital age, concerns about online security and privacy have become paramount. However, addressing these issues can be difficult, especially within the context of family relationships, wherein parents and children may have conflicting interests. In this environment, parents and children may turn to online security advice to determine how to proceed. In this paper, we examine the advice available to parents and children regarding content filtering and circumvention as found on YouTube and TikTok. In an analysis of 839 videos returned from queries on these topics, we found that half (n=399) provide relevant advice. Our results show that of these videos, roughly three-quarters are accurate, with the remaining one-fourth containing factually incorrect advice. We find that videos targeting children are both more likely to be incorrect and actionable than videos targeting parents, leaving children at increased risk of taking harmful action. Moreover, we find that while advice videos targeting parents will occasionally discuss the ethics of content filtering and device monitoring (including recommendations to respect children's autonomy) no such discussion of the ethics or risks of circumventing content filtering is given to children, leaving them unaware of any risks that may be involved with doing so. Ultimately, our research indicates that video-based social media sites are already effective sources of security advice propagation and that the public would benefit from security researchers and practitioners engaging more with these platforms, both for the creation of content and of tools designed to help with more effective filtering., Comment: 15 pages, 5 figures, 8 tables

Quantization is a fundamental optimization for many machine-learning use cases, including compressing gradients, model weights and activations, and datasets. The most accurate form of quantization is \emph{adaptive}, where the error is minimized with respect to a given input, rather than optimizing for the worst case. However, optimal adaptive quantization methods are considered infeasible in terms of both their runtime and memory requirements. We revisit the Adaptive Vector Quantization (AVQ) problem and present algorithms that find optimal solutions with asymptotically improved time and space complexity. We also present an even faster near-optimal algorithm for large inputs. Our experiments show our algorithms may open the door to using AVQ more extensively in a variety of machine learning applications.

Gradient inversion attacks aim to reconstruct local training data from intermediate gradients exposed in the federated learning framework. Despite successful attacks, all previous methods, starting from reconstructing a single data point and then relaxing the single-image limit to batch level, are only tested under hard label constraints. Even for single-image reconstruction, we still lack an analysis-based algorithm to recover augmented soft labels. In this work, we change the focus from enlarging batchsize to investigating the hard label constraints, considering a more realistic circumstance where label smoothing and mixup techniques are used in the training process. In particular, we are the first to initiate a novel algorithm to simultaneously recover the ground-truth augmented label and the input feature of the last fully-connected layer from single-input gradients, and provide a necessary condition for any analytical-based label recovery methods. Extensive experiments testify to the label recovery accuracy, as well as the benefits to the following image reconstruction. We believe soft labels in classification tasks are worth further attention in gradient inversion attacks., Comment: ICLR2024 poster

Patent documents in the patent database (PatDB) are crucial for research, development, and innovation as they contain valuable technical information. However, PatDB presents a multifaceted challenge compared to publicly available preprocessed databases due to the intricate nature of the patent text and the inherent sparsity within the patent citation network. Although patent text analysis and citation analysis bring new opportunities to explore patent data mining, no existing work exploits the complementation of them. To this end, we propose a joint semantic-topological evolutionary graph learning approach (PatSTEG) to model the formation dynamics of patent citation networks. More specifically, we first create a real-world dataset of Chinese patents named CNPat and leverage its patent texts and citations to construct a patent citation network. Then, PatSTEG is modeled to study the evolutionary dynamics of patent citation formation by considering the semantic and topological information jointly. Extensive experiments are conducted on CNPat and public datasets to prove the superiority of PatSTEG over other state-of-the-art methods. All the results provide valuable references for patent literature research and technical exploration.

$ $Large Language Models (LLMs) are being increasingly utilized in various applications, with code generations being a notable example. While previous research has shown that LLMs have the capability to generate both secure and insecure code, the literature does not take into account what factors help generate secure and effective code. Therefore in this paper we focus on identifying and understanding the conditions and contexts in which LLMs can be effectively and safely deployed in real-world scenarios to generate quality code. We conducted a comparative analysis of four advanced LLMs--GPT-3.5 and GPT-4 using ChatGPT and Bard and Gemini from Google--using 9 separate tasks to assess each model's code generation capabilities. We contextualized our study to represent the typical use cases of a real-life developer employing LLMs for everyday tasks as work. Additionally, we place an emphasis on security awareness which is represented through the use of two distinct versions of our developer persona. In total, we collected 61 code outputs and analyzed them across several aspects: functionality, security, performance, complexity, and reliability. These insights are crucial for understanding the models' capabilities and limitations, guiding future development and practical applications in the field of automated code generation., Comment: 12 pages, 2 figures

Most vision-based tactile sensors use elastomer deformation to infer tactile information, which can not sense some modalities, like temperature. As an important part of human tactile perception, temperature sensing can help robots better interact with the environment. In this work, we propose a novel multimodal vision-based tactile sensor, SATac, which can simultaneously perceive information of temperature, pressure, and shear. SATac utilizes thermoluminescence of strontium aluminate (SA) to sense a wide range of temperatures with exceptional resolution. Additionally, the pressure and shear can also be perceived by analyzing Voronoi diagram. A series of experiments are conducted to verify the performance of our proposed sensor. We also discuss the possible application scenarios and demonstrate how SATac could benefit robot perception capabilities.

We consider the dispersion managed nonlinear Schr\"dinger equations with quintic and cubic nonlinearities in one and two dimensions, respectively. We prove the global well-posedness and scattering in $L_x^2$ for small initial data employing the $U^p$ and $V^p$ spaces., Comment: 14 pages

Patterns by self-organization in nature have garnered significant interest in a range of disciplines due to their intriguing structures. In the context of the snowdrift game (SDG), which is considered as an anti-coordination game, but the anti-coordination patterns are counterintuitively rare. In the work, we introduce a model called the Two-Agents, Two-Action Reinforcement Learning Evolutionary Game ($2\times 2$ RLEG), and apply it to the SDG on regular lattices. We uncover intriguing phenomena in the form of Anti-Coordinated domains (AC-domains), where different frustration regions are observed and continuous phase transitions at the boundaries are identified. To understand the underlying mechanism, we develop a perturbation theory to analyze the stability of different AC-domains. Our theory accurately partitions the parameter space into non-anti-coordinated, anti-coordinated, and mixed areas, and captures their dependence on the learning parameters. Lastly, abnormal scenarios with a large learning rate and a large discount factor that deviate from the theory are investigated by examining the growth and nucleation of AC-domains. Our work provides insights into the emergence of spatial patterns in nature, and contributes to the development of theory for analysing their structural complexities.

We consider the Gabitov-Turitsyn equation or the dispersion managed nonlinear Schr\"odinger equation of a power-type nonlinearity \[ i\partial_t u+ d_\text{av} \partial_x^2u+\int_0^1 e^{-ir\partial_x^2}\big(|e^{ir\partial_x^2}u|^{p-1}e^{ir\partial_x^2}u\big)dr=0 \] and prove the global existence versus finite time blowup dichotomy for the mass-supercritical cases, that is, $p>9$., Comment: 23pages. 1 figure

To investigate the influence of inertia and slip on the instability of a liquid film on a fibre, a theoretical framework based on the axisymmetric Navier-Stokes equations is proposed via linear instability analysis. The model reveals that slip significantly enhances perturbation growth in viscous film flows, whereas it exerts minimal influence on flows dominated by inertia. Moreover, under no-slip boundary conditions, the dominant instability mode of thin films remains unaltered by inertia, closely aligning with predictions from a no-slip lubrication model. Conversely, when slip is introduced, the dominant wavenumber experiences a noticeable reduction as inertia decreases. This trend is captured by an introduced lubrication model with giant slip. Direct numerical simulations of the Navier-Stokes equations are then performed to further confirm the theoretical findings at the linear stage. For the nonlinear dynamics, no-slip simulations show complex vortical structures within films, driven by fluid inertia near surfaces. Additionally, in scenarios with weak inertia, a reduction in the volume of satellite droplets is observed due to slip, following a power-law relationship.

Constructing models to discover physics underlying magnanimous data is a traditional strategy in data mining which has been proved to be powerful and successful. In this work, a multi-optimized recurrent neural network (MRNN) is utilized to predict the dynamics of photosynthetic excitation energy transfer (EET) in a light-harvesting complex. The original data set produced by the master equation were trained to forecast the EET evolution. An agreement between our prediction and the theoretical deduction with an accuracy of over 99.26\% is found, showing the validity of the proposed MRNN. A time-segment polynomial fitting multiplied by a unit step function results in a loss rate of the order of $10^{-5}$, showing a striking consistence with analytical formulations for the photosynthetic EET. The work sets up a precedent for accurate EET prediction from large data set by establishing analytical descriptions for physics hidden behind, through minimizing the processing cost during the evolution of week-coupling EET., Comment: 16 pages, 9 figures

Here we present the training and evaluation of NanoNER, a Named Entity Recognition (NER) model for Nanobiology. NER consists in the identification of specific entities in spans of unstructured texts and is often a primary task in Natural Language Processing (NLP) and Information Extraction. The aim of our model is to recognise entities previously identified by domain experts as constituting the essential knowledge of the domain. Relying on ontologies, which provide us with a domain vocabulary and taxonomy, we implemented an iterative process enabling experts to determine the entities relevant to the domain at hand. We then delve into the potential of distant supervision learning in NER, supporting how this method can increase the quantity of annotated data with minimal additional manpower. On our full corpus of 728 full-text nanobiology articles, containing more than 120k entity occurrences, NanoNER obtained a F1-score of 0.98 on the recognition of previously known entities. Our model also demonstrated its ability to discover new entities in the text, with precision scores ranging from 0.77 to 0.81. Ablation experiments further confirmed this and allowed us to assess the dependency of our approach on the external resources. It highlighted the dependency of the approach to the resource, while also confirming its ability to rediscover up to 30% of the ablated terms. This paper details the methodology employed, experimental design, and key findings, providing valuable insights and directions for future related researches on NER in specialized domain. Furthermore, since our approach require minimal manpower , we believe that it can be generalized to other specialized fields.

The duration of the accumulation rate (physical image) is a key factor in analysis of counterintuitive phenomena involving relative timescales on multiplex networks. Typically, the relative timescales are represented by multiplying any layer by the same factor. However, researchers often overlook the changes in the relative timescales caused by local parameters, resulting in incomplete analysis of phenomena. This paper examines the survival time of stifler individuals in the information-epidemic model on multiplex networks. The relative timescales can be affected by the survival time (only one parameter), reversing the monotonically increasing phenomenon into a monotonically decreasing one, that is, a counterintuitive phenomenon under incomplete analysis. Additionally, the relative timescales can influence the epidemic threshold, which is different from the previous studies. Our work suggests that considering the physical image of relative timescales is crucial when analyzing multiplex networks, even when only one parameter is altered.

Few-Shot Class-Incremental Learning (FSCIL) aims to enable deep neural networks to learn new tasks incrementally from a small number of labeled samples without forgetting previously learned tasks, closely mimicking human learning patterns. In this paper, we propose a novel approach called Prompt Learning for FSCIL (PL-FSCIL), which harnesses the power of prompts in conjunction with a pre-trained Vision Transformer (ViT) model to address the challenges of FSCIL effectively. Our work pioneers the use of visual prompts in FSCIL, which is characterized by its notable simplicity. PL-FSCIL consists of two distinct prompts: the Domain Prompt and the FSCIL Prompt. Both are vectors that augment the model by embedding themselves into the attention layer of the ViT model. Specifically, the Domain Prompt assists the ViT model in adapting to new data domains. The task-specific FSCIL Prompt, coupled with a prototype classifier, amplifies the model's ability to effectively handle FSCIL tasks. We validate the efficacy of PL-FSCIL on widely used benchmark datasets such as CIFAR-100 and CUB-200. The results showcase competitive performance, underscoring its promising potential for real-world applications where high-quality data is often scarce. The source code is available at: https://github.com/TianSongS/PL-FSCIL.

We studied the Hayden-Preskill thought experiment with the local projective measurement. Compared to the original model, the measurement is applied on the Hawking radiation that was emitted after throwing the quantum diary into the black hole. Within this setup, we explored various aspects of this model, including the information recovery from the black hole, the relation to the black hole final state proposal, the relation between the Yoshida-Kitaev protocol and Petz recovery map, the effects of the decoherence, and the quantum simulations of the decoding protocols. These aspects may provide us new insights into the non-perturbative nature of quantum black holes.

Electrical manipulation of magnetic order by current-induced spin torques lays the foundation for spintronics. One promising approach is encoding information in the N\'eel vector of antiferromagnetic (AFM) materials, particularly to collinear antiferromagnets with the perpendicular magnetic anisotropy (PMA), as the negligible stray fields and terahertz spin dynamics can enable memory devices with higher integration density and ultrafast speed. Here we demonstrate that the N\'eel order information in a prototypical collinear AFM insulator with PMA, Cr2O3, can be reliably readout via the anomalous Hall effect and efficiently switched by the spin-orbit torque (SOT) effect with a low current density of 5.8*106 A/cm2. Moreover, using Cr2O3 as a mediator, we electrically switch the magnetization of a Y3Fe5O12 film exchange-coupled to the Cr2O3 layer, unambiguously confirming the N\'eel order switching of the Cr2O3 layer. This work provides a significant basis for developing AFM memory devices based on collinear AFM materials with PMA.

The classification of carotid artery ultrasound images is a crucial means for diagnosing carotid plaques, holding significant clinical relevance for predicting the risk of stroke. Recent research suggests that utilizing plaque segmentation as an auxiliary task for classification can enhance performance by leveraging the correlation between segmentation and classification tasks. However, this approach relies on obtaining a substantial amount of challenging-to-acquire segmentation annotations. This paper proposes a novel weakly supervised auxiliary task learning network model (WAL-Net) to explore the interdependence between carotid plaque classification and segmentation tasks. The plaque classification task is primary task, while the plaque segmentation task serves as an auxiliary task, providing valuable information to enhance the performance of the primary task. Weakly supervised learning is adopted in the auxiliary task to completely break away from the dependence on segmentation annotations. Experiments and evaluations are conducted on a dataset comprising 1270 carotid plaque ultrasound images from Wuhan University Zhongnan Hospital. Results indicate that the proposed method achieved an approximately 1.3% improvement in carotid plaque classification accuracy compared to the baseline network. Specifically, the accuracy of mixed-echoic plaques classification increased by approximately 3.3%, demonstrating the effectiveness of our approach.

Ensuring the accuracy of responses provided by large language models (LLMs) is crucial, particularly in clinical settings where incorrect information may directly impact patient health. To address this challenge, we construct K-QA, a dataset containing 1,212 patient questions originating from real-world conversations held on K Health (an AI-driven clinical platform). We employ a panel of in-house physicians to answer and manually decompose a subset of K-QA into self-contained statements. Additionally, we formulate two NLI-based evaluation metrics approximating recall and precision: (1) comprehensiveness, measuring the percentage of essential clinical information in the generated answer and (2) hallucination rate, measuring the number of statements from the physician-curated response contradicted by the LLM answer. Finally, we use K-QA along with these metrics to evaluate several state-of-the-art models, as well as the effect of in-context learning and medically-oriented augmented retrieval schemes developed by the authors. Our findings indicate that in-context learning improves the comprehensiveness of the models, and augmented retrieval is effective in reducing hallucinations. We make K-QA available to to the community to spur research into medically accurate NLP applications., Comment: The data and the evaluation script are available at https://github.com/Itaymanes/K-QA. Results and model comparisons can be viewed at https://huggingface.co/spaces/Itaykhealth/K-QA

We report quantum oscillation measurements on the kagome compounds ATi$_3$Bi$_5$ (A=Rb, Cs) in magnetic fields up to 41.5 T and temperatures down to 350 mK. In addition to the frequencies observed in previous studies, we have observed multiple unreported frequencies above 2000 T in CsTi$_3$Bi$_5$ using a tunnel diode oscillator technique. We compare these results against density functional theory calculations and find good agreement with the calculations in the number of peaks observed, frequency, and the dimensionality of the Fermi surface. For RbTi$_3$Bi$_5$ we have obtained a different quantum oscillation spectrum, although calculated quantum oscillation frequencies for the Rb compound are remarkably similar to the Cs compound, calling for further studies.

The field of healthcare has increasingly turned its focus towards Large Language Models (LLMs) due to their remarkable performance. However, their performance in actual clinical applications has been underexplored. Traditional evaluations based on question-answering tasks don't fully capture the nuanced contexts. This gap highlights the need for more in-depth and practical assessments of LLMs in real-world healthcare settings. Objective: We sought to evaluate the performance of LLMs in the complex clinical context of adult critical care medicine using systematic and comprehensible analytic methods, including clinician annotation and adjudication. Methods: We investigated the performance of three general LLMs in understanding and processing real-world clinical notes. Concepts from 150 clinical notes were identified by MetaMap and then labeled by 9 clinicians. Each LLM's proficiency was evaluated by identifying the temporality and negation of these concepts using different prompts for an in-depth analysis. Results: GPT-4 showed overall superior performance compared to other LLMs. In contrast, both GPT-3.5 and text-davinci-003 exhibit enhanced performance when the appropriate prompting strategies are employed. The GPT family models have demonstrated considerable efficiency, evidenced by their cost-effectiveness and time-saving capabilities. Conclusion: A comprehensive qualitative performance evaluation framework for LLMs is developed and operationalized. This framework goes beyond singular performance aspects. With expert annotations, this methodology not only validates LLMs' capabilities in processing complex medical data but also establishes a benchmark for future LLM evaluations across specialized domains.

Intelligent robot is the ultimate goal in the robotics field. Existing works leverage learning-based or optimization-based methods to accomplish human-defined tasks. However, the challenge of enabling robots to explore various environments autonomously remains unresolved. In this work, we propose a framework named GExp, which enables robots to explore and learn autonomously without human intervention. To achieve this goal, we devise modules including self-exploration, knowledge-base-building, and close-loop feedback based on foundation models. Inspired by the way that infants interact with the world, GExp encourages robots to understand and explore the environment with a series of self-generated tasks. During the process of exploration, the robot will acquire skills from beneficial experiences that are useful in the future. GExp provides robots with the ability to solve complex tasks through self-exploration. GExp work is independent of prior interactive knowledge and human intervention, allowing it to adapt directly to different scenarios, unlike previous studies that provided in-context examples as few-shot learning. In addition, we propose a workflow of deploying the real-world robot system with self-learned skills as an embodied assistant., Comment: 19 pages

Higher-order nodal line semimetals represent a recently proposed topological semimetal class that harbors bulk nodal lines and features gapless hinge Fermi arc excitations, governed by the bulk-hinge correspondence. In this study, we investigate the disorder effect on a higher-order nodal line semimetal and the consequent phase transitions. Within the pristine higher-order nodal line semimetal model, we unveil three distinct phases: higher-order nodal line semimetal, conventional nodal line semimetal, and normal insulator. The higher-order nodal line semimetal is characterized by one-dimensional hinge Fermi arc states connecting a pair of nodal rings, contrasting with conventional nodal line semimetals that exhibit two-dimensional drumhead surface states. We demonstrate that disorder can trigger multiple phase transitions within this system. Significantly, intermediate disorder can induce higher-order topology in an initial conventional nodal line semimetal or even an initial normal insulator. Further increase in disorder drives the system through a diffusive metallic phase before ultimately reaching the Anderson insulator regime. Employing a combination of finite-size scaling analysis and an effective medium theory, we construct a comprehensive phase diagram, elucidating the intricate interplay between disorder and topology., Comment: 8 pages, 9 figures.Comments are welcome

We examine the nonequilibrium nature of two-phase fluid displacements in a quasi-two-dimensional medium (a model open fracture), in the presence of localized constrictions ("defects"), from a theoretical and numerical standpoint. Our analysis predicts the capillary energy dissipated in abrupt interfacial displacements (jumps) across defects, and relates it to the corresponding hysteresis cycle, e.g. in pressure-saturation. We distinguish between "weak" (reversible interface displacement, exhibiting no hysteresis and dissipation) and "strong" (irreversible) defects. We expose the emergence of dissipation and irreversibility caused by spatial interactions, mediated by interfacial tension, among otherwise weak defects. We exemplify this cooperative behavior for a pair of weak defects and establish a critical separation distance, analytically and numerically, verified by a proof-of-concept experiment.

Truck parking on freight corridors faces various challenges, such as insufficient parking spaces and compliance with Hour-of-Service (HOS) regulations. These constraints often result in unauthorized parking practices, causing safety concerns. To enhance the safety of freight operations, providing accurate parking usage prediction proves to be a cost-effective solution. Despite the existing research demonstrating satisfactory accuracy for predicting individual truck parking site usage, few approaches have been proposed for predicting usage with spatial dependencies of multiple truck parking sites. We present the Regional Temporal Graph Neural Network (RegT-GCN) as a predictive framework for assessing parking usage across the entire state to provide better truck parking information and mitigate unauthorized parking. The framework leverages the topological structures of truck parking site distributions and historical parking data to predict occupancy rates across a state. To achieve this, we introduce a Regional Decomposition approach, which effectively captures the geographical characteristics. We also introduce the spatial module working efficiently with the temporal module. Evaluation results demonstrate that the proposed model surpasses other baseline models, improving the performance by more than $20\%$ compared with the original model. The proposed model allows truck parking sites' percipience of the topological structures and provides higher performance., Comment: 10 pages, 5 figures, 3 tables, Manuscript for IEEE Transactions on Intelligent Transportation Systems

We report 850 $\mu$m continuum polarization observations toward the filamentary high-mass star-forming region NGC 2264, taken as part of the B-fields In STar forming Regions Observations (BISTRO) large program on the James Clerk Maxwell Telescope (JCMT). These data reveal a well-structured non-uniform magnetic field in the NGC 2264C and 2264D regions with a prevailing orientation around 30 deg from north to east. Field strengths estimates and a virial analysis for the major clumps indicate that NGC 2264C is globally dominated by gravity while in 2264D magnetic, gravitational, and kinetic energies are roughly balanced. We present an analysis scheme that utilizes the locally resolved magnetic field structures, together with the locally measured gravitational vector field and the extracted filamentary network. From this, we infer statistical trends showing that this network consists of two main groups of filaments oriented approximately perpendicular to one another. Additionally, gravity shows one dominating converging direction that is roughly perpendicular to one of the filament orientations, which is suggestive of mass accretion along this direction. Beyond these statistical trends, we identify two types of filaments. The type-I filament is perpendicular to the magnetic field with local gravity transitioning from parallel to perpendicular to the magnetic field from the outside to the filament ridge. The type-II filament is parallel to the magnetic field and local gravity. We interpret these two types of filaments as originating from the competition between radial collapsing, driven by filament self-gravity, and the longitudinal collapsing, driven by the region's global gravity., Comment: Accepted for publication in the Astrophysical Journal. 43 pages, 32 figures, and 4 tables (including Appendix)

In this study, we construct a series of evolving epidemic networks by measuring the correlations of daily COVID-19 cases time series among 3,105 counties in the United States. Remarkably, through quantitative analysis of the spatial distribution of these entities in different networks, we identify four typical patterns of COVID-19 transmission in the United States from March 2020 to February 2023. The onsets and wanes of these patterns are closely associated with significant events in the COVID-19 timeline. Furthermore, we conduct in-depth qualitative and quantitative research on the spread of the epidemic at the county and state levels, tracing and analyzing the evolution and characteristics of specific propagation pathways. Overall, our research breaks away from traditional infectious disease models and provides a macroscopic perspective on the evolution in epidemic transmission patterns. This highlights the remarkable potential of utilizing complex network methods for macroscopic studies of infectious diseases., Comment: 13 pages, 5 figures. This is a preprint of the final version

Recently, formal verification of deep neural networks (DNNs) has garnered considerable attention, and over-approximation based methods have become popular due to their effectiveness and efficiency. However, these strategies face challenges in addressing the "unknown dilemma" concerning whether the exact output region or the introduced approximation error violates the property in question. To address this, this paper introduces the UR4NNV verification framework, which utilizes under-approximation reachability analysis for DNN verification for the first time. UR4NNV focuses on DNNs with Rectified Linear Unit (ReLU) activations and employs a binary tree branch-based under-approximation algorithm. In each epoch, UR4NNV under-approximates a sub-polytope of the reachable set and verifies this polytope against the given property. Through a trial-and-error approach, UR4NNV effectively falsifies DNN properties while providing confidence levels when reaching verification epoch bounds and failing falsifying properties. Experimental comparisons with existing verification methods demonstrate the effectiveness and efficiency of UR4NNV, significantly reducing the impact of the "unknown dilemma"., Comment: 11 pages, 4 figures