Search

Artificial intelligence has, so far, largely automated routine tasks, but what does it mean for the future of work if Large Language Models (LLMs) show creativity comparable to humans? To measure the creativity of LLMs holistically, the current study uses 13 creative tasks spanning three domains. We benchmark the LLMs against individual humans, and also take a novel approach by comparing them to the collective creativity of groups of humans. We find that the best LLMs (Claude and GPT-4) rank in the 52nd percentile against humans, and overall LLMs excel in divergent thinking and problem solving but lag in creative writing. When questioned 10 times, an LLM's collective creativity is equivalent to 8-10 humans. When more responses are requested, two additional responses of LLMs equal one extra human. Ultimately, LLMs, when optimally applied, may compete with a small group of humans in the future of work.

A 2D electromagnetic-thermal coupled numerical model has been developed using the finite element method and validated against experimental data to investigate a superconducting machine featuring high-temperature superconducting (HTS) tape stacks and cryocooled iron cores. The HTS stacks are transformed into trapped field stacks (TFSs) through pulsed field magnetisation (PFM), generating rotor fields. After PFM, the superconducting motor operates on the same principle as permanent magnet synchronous motors. This study explores the behaviour of HTS stacks by altering the stack's layer number from one to nine and adjusting the pulsed current amplitude from 250 A to 1000 A. The primary objective of this paper is to identify the optimal combination of pulsed current amplitudes and TFS layer numbers for achieving maximum magnetisation fields. The secondary objective is to evaluate the overall losses in both superconducting and non-superconducting parts of the machine during magnetisation, including heat generated in various layers of the TFS, and losses in the motor's active materials (copper windings and iron cores). Two motor configurations were proposed, and two calculation methods using linear interpolation of iron losses and steel grades were introduced to estimate the iron losses for the studied iron material, M270-35A. This pioneering study is expected to serve as a valuable reference for loss analyses and structural design considerations in developing superconducting machines., Comment: 19pages, 18 figures, 5 tables

Large Language Models (LLMs) have been widely adopted to process long-context tasks. However, the large memory overhead of the key-value (KV) cache poses significant challenges in long-context scenarios. Existing training-free KV cache compression methods typically focus on quantization and token pruning, which have compression limits, and excessive sparsity can lead to severe performance degradation. Other methods design new architectures with less KV overhead but require significant training overhead. To address the above two drawbacks, we further explore the redundancy in the channel dimension and apply an architecture-level design with minor training costs. Therefore, we introduce CSKV, a training-efficient Channel Shrinking technique for KV cache compression: (1) We first analyze the singular value distribution of the KV cache, revealing significant redundancy and compression potential along the channel dimension. Based on this observation, we propose using low-rank decomposition for key and value layers and storing the low-dimension features. (2) To preserve model performance, we introduce a bi-branch KV cache, including a window-based full-precision KV cache and a low-precision compressed KV cache. (3) To reduce the training costs, we minimize the layer-wise reconstruction loss for the compressed KV cache instead of retraining the entire LLMs. Extensive experiments show that CSKV can reduce the memory overhead of the KV cache by 80% while maintaining the model's long-context capability. Moreover, we show that our method can be seamlessly combined with quantization to further reduce the memory overhead, achieving a compression ratio of up to 95%. Code is available at https://github.com/wln20/CSKV., Comment: 4th NeurIPS Efficient Natural Language and Speech Processing Workshop (ENLSP-IV 2024)

Deep neural networks (DNNs) have significantly boosted the performance of many challenging tasks. Despite the great development, DNNs have also exposed their vulnerability. Recent studies have shown that adversaries can manipulate the predictions of DNNs by adding a universal adversarial perturbation (UAP) to benign samples. On the other hand, increasing efforts have been made to help users understand and explain the inner working of DNNs by highlighting the most informative parts (i.e., attribution maps) of samples with respect to their predictions. Moreover, we first empirically find that such attribution maps between benign and adversarial examples have a significant discrepancy, which has the potential to detect universal adversarial perturbations for defending against adversarial attacks. This finding motivates us to further investigate a new research problem: whether there exist universal adversarial perturbations that are able to jointly attack DNNs classifier and its interpretation with malicious desires. It is challenging to give an explicit answer since these two objectives are seemingly conflicting. In this paper, we propose a novel attacking framework to generate joint universal adversarial perturbations (JUAP), which can fool the DNNs model and misguide the inspection from interpreters simultaneously. Comprehensive experiments on various datasets demonstrate the effectiveness of the proposed method JUAP for joint attacks. To the best of our knowledge, this is the first effort to study UAP for jointly attacking both DNNs and interpretations.

Large Language Models (LLMs) have attracted extensive attention due to their remarkable performance across various tasks. However, the substantial computational and memory requirements of LLM inference pose challenges for deployment in resource-constrained scenarios. Efforts within the field have been directed towards developing techniques aimed at enhancing the efficiency of LLM inference. This paper presents a comprehensive survey of the existing literature on efficient LLM inference. We start by analyzing the primary causes of the inefficient LLM inference, i.e., the large model size, the quadratic-complexity attention operation, and the auto-regressive decoding approach. Then, we introduce a comprehensive taxonomy that organizes the current literature into data-level, model-level, and system-level optimization. Moreover, the paper includes comparative experiments on representative methods within critical sub-fields to provide quantitative insights. Last but not least, we provide some knowledge summary and discuss future research directions.

We report results obtained and insights gained while answering the following question: how effective is it to use a simulator to establish path following control policies for an autonomous ground robot? While the quality of the simulator conditions the answer to this question, we found that for the simulation platform used herein, producing four control policies for path planning was straightforward once a digital twin of the controlled robot was available. The control policies established in simulation and subsequently demonstrated in the real world are PID control, MPC, and two neural network (NN) based controllers. Training the two NN controllers via imitation learning was accomplished expeditiously using seven simple maneuvers: follow three circles clockwise, follow the same circles counter-clockwise, and drive straight. A test randomization process that employs random micro-simulations is used to rank the ``goodness'' of the four control policies. The policy ranking noted in simulation correlates well with the ranking observed when the control policies were tested in the real world. The simulation platform used is publicly available and BSD3-released as open source; a public Docker image is available for reproducibility studies. It contains a dynamics engine, a sensor simulator, a ROS2 bridge, and a ROS2 autonomy stack the latter employed both in the simulator and the real world experiments., Comment: 8 pages, 7 figures

Post-training quantization (PTQ) has emerged as a promising technique to reduce the cost of large language models (LLMs). Specifically, PTQ can effectively mitigate memory consumption and reduce computational overhead in LLMs. To meet the requirements of both high efficiency and performance across diverse scenarios, a comprehensive evaluation of quantized LLMs is essential to guide the selection of quantization methods. This paper presents a thorough evaluation of these factors by evaluating the effect of PTQ on Weight, Activation, and KV Cache on 11 model families, including OPT, LLaMA2, Falcon, Bloomz, Mistral, ChatGLM, Vicuna, LongChat, StableLM, Gemma, and Mamba, with parameters ranging from 125M to 180B. The evaluation encompasses five types of tasks: basic NLP, emergent ability, trustworthiness, dialogue, and long-context tasks. Moreover, we also evaluate the state-of-the-art (SOTA) quantization methods to demonstrate their applicability. Based on the extensive experiments, we systematically summarize the effect of quantization, provide recommendations to apply quantization techniques, and point out future directions. The code can be found in https://github.com/thu-nics/qllm-eval.

The QED-C suite of Application-Oriented Benchmarks provides the ability to gauge performance characteristics of quantum computers as applied to real-world applications. Its benchmark programs sweep over a range of problem sizes and inputs, capturing key performance metrics related to the quality of results, total time of execution, and quantum gate resources consumed. In this manuscript, we investigate challenges in broadening the relevance of this benchmarking methodology to applications of greater complexity. First, we introduce a method for improving landscape coverage by varying algorithm parameters systematically, exemplifying this functionality in a new scalable HHL linear equation solver benchmark. Second, we add a VQE implementation of a Hydrogen Lattice simulation to the QED-C suite, and introduce a methodology for analyzing the result quality and run-time cost trade-off. We observe a decrease in accuracy with increased number of qubits, but only a mild increase in the execution time. Third, unique characteristics of a supervised machine-learning classification application are explored as a benchmark to gauge the extensibility of the framework to new classes of application. Applying this to a binary classification problem revealed the increase in training time required for larger anzatz circuits, and the significant classical overhead. Fourth, we add methods to include optimization and error mitigation in the benchmarking workflow which allows us to: identify a favourable trade off between approximate gate synthesis and gate noise; observe the benefits of measurement error mitigation and a form of deterministic error mitigation algorithm; and to contrast the improvement with the resulting time overhead. Looking ahead, we discuss how the benchmark framework can be instrumental in facilitating the exploration of algorithmic options and their impact on performance., Comment: 21 pages, 21 figures

The electron pair approximation offers a resource efficient variational quantum eigensolver (VQE) approach for quantum chemistry simulations on quantum computers. With the number of entangling gates scaling quadratically with system size and a constant energy measurement overhead, the orbital optimized unitary pair coupled cluster double (oo-upCCD) ansatz strikes a balance between accuracy and efficiency on today's quantum computers. However, the electron pair approximation makes the method incapable of producing quantitatively accurate energy predictions. In order to improve the accuracy without increasing the circuit depth, we explore the idea of reduced density matrix (RDM) based second order perturbation theory (PT2) as an energetic correction to electron pair approximation. The new approach takes into account of the broken-pair energy contribution that is missing in pair-correlated electron simulations, while maintaining the computational advantages of oo-upCCD ansatz. In dissociations of N$_2$, Li$_2$O, and chemical reactions such as the unimolecular decomposition of CH$_2$OH$^+$ and the \snTwo reaction of CH$_3$I $+$ Br$^-$, the method significantly improves the accuracy of energy prediction. On two generations of the IonQ's trapped ion quantum computers, Aria and Forte, we find that unlike the VQE energy, the PT2 energy correction is highly noise-resilient. By applying a simple error mitigation approach based on post-selection solely on the VQE energies, the predicted VQE-PT2 energy differences between reactants, transition state, and products are in excellent agreement with noise-free simulators.

Turbulent flows have historically presented formidable challenges to predictive computational modeling. Traditional numerical simulations often require vast computational resources, making them infeasible for numerous engineering applications. As an alternative, deep learning-based surrogate models have emerged, offering data-drive solutions. However, these are typically constructed within deterministic settings, leading to shortfall in capturing the innate chaotic and stochastic behaviors of turbulent dynamics. We introduce a novel generative framework grounded in probabilistic diffusion models for versatile generation of spatiotemporal turbulence. Our method unifies both unconditional and conditional sampling strategies within a Bayesian framework, which can accommodate diverse conditioning scenarios, including those with a direct differentiable link between specified conditions and generated unsteady flow outcomes, and scenarios lacking such explicit correlations. A notable feature of our approach is the method proposed for long-span flow sequence generation, which is based on autoregressive gradient-based conditional sampling, eliminating the need for cumbersome retraining processes. We showcase the versatile turbulence generation capability of our framework through a suite of numerical experiments, including: 1) the synthesis of LES simulated instantaneous flow sequences from URANS inputs; 2) holistic generation of inhomogeneous, anisotropic wall-bounded turbulence, whether from given initial conditions, prescribed turbulence statistics, or entirely from scratch; 3) super-resolved generation of high-speed turbulent boundary layer flows from low-resolution data across a range of input resolutions. Collectively, our numerical experiments highlight the merit and transformative potential of the proposed methods, making a significant advance in the field of turbulence generation., Comment: 37 pages, 31 figures

This paper introduces DEM-Engine, a new submodule of Project Chrono, that is designed to carry out Discrete Element Method (DEM) simulations. Based on spherical primitive shapes, DEM-Engine can simulate polydisperse granular materials and handle complex shapes generated as assemblies of primitives, referred to as clumps. DEM-Engine has a multi-tier parallelized structure that is optimized to operate simultaneously on two GPUs. The code uses custom-defined data types to reduce memory footprint and increase bandwidth. A novel "delayed contact detection" algorithm allows the decoupling of the contact detection and force computation, thus splitting the workload into two asynchronous GPU streams. DEM-Engine uses just-in-time compilation to support user-defined contact force models. This paper discusses its C++ and Python interfaces and presents a variety of numerical tests, in which impact forces, complex-shaped particle flows, and a custom force model are validated considering well-known benchmark cases. Additionally, the full potential of the simulator is demonstrated for the investigation of extraterrestrial rover mobility on granular terrain. The chosen case study demonstrates that large-scale co-simulations (comprising 11 million elements) spanning 15 seconds, in conjunction with an external multi-body dynamics system, can be efficiently executed within a day. Lastly, a performance test suggests that DEM-Engine displays linear scaling up to 150 million elements on two NVIDIA A100 GPUs., Comment: 38 pages, 30 figures, 9 tables. This preprint is submitted to Computer Physics Communications

Accurate and scalable methods for computational quantum chemistry can accelerate research and development in many fields, ranging from drug discovery to advanced material design. Solving the electronic Schrodinger equation is the core problem of computational chemistry. However, the combinatorial complexity of this problem makes it intractable to find exact solutions, except for very small systems. The idea of quantum computing originated from this computational challenge in simulating quantum-mechanics. We propose an end-to-end quantum chemistry pipeline based on the variational quantum eigensolver (VQE) algorithm and integrated with both HPC-based simulators and a trapped-ion quantum computer. Our platform orchestrates hundreds of simulation jobs on compute resources to efficiently complete a set of ab initio chemistry experiments with a wide range of parameterization. Per- and poly-fluoroalkyl substances (PFAS) are a large family of human-made chemicals that pose a major environmental and health issue globally. Our simulations includes breaking a Carbon-Fluorine bond in trifluoroacetic acid (TFA), a common PFAS chemical. This is a common pathway towards destruction and removal of PFAS. Molecules are modeled on both a quantum simulator and a trapped-ion quantum computer, specifically IonQ Aria. Using basic error mitigation techniques, the 11-qubit TFA model (56 entangling gates) on IonQ Aria yields near-quantitative results with milli-Hartree accuracy. Our novel results show the current state and future projections for quantum computing in solving the electronic structure problem, push the boundaries for the VQE algorithm and quantum computers, and facilitates development of quantum chemistry workflows.

Comprehensive and accurate evaluation of general-purpose AI systems such as large language models allows for effective mitigation of their risks and deepened understanding of their capabilities. Current evaluation methodology, mostly based on benchmarks of specific tasks, falls short of adequately assessing these versatile AI systems, as present techniques lack a scientific foundation for predicting their performance on unforeseen tasks and explaining their varying performance on specific task items or user inputs. Moreover, existing benchmarks of specific tasks raise growing concerns about their reliability and validity. To tackle these challenges, we suggest transitioning from task-oriented evaluation to construct-oriented evaluation. Psychometrics, the science of psychological measurement, provides a rigorous methodology for identifying and measuring the latent constructs that underlie performance across multiple tasks. We discuss its merits, warn against potential pitfalls, and propose a framework to put it into practice. Finally, we explore future opportunities of integrating psychometrics with the evaluation of general-purpose AI systems.

We report on a study that employs an in-house developed simulation infrastructure to accomplish zero shot policy transferability for a control policy associated with a scale autonomous vehicle. We focus on implementing policies that require no real world data to be trained (Zero-Shot Transfer), and are developed in-house as opposed to being validated by previous works. We do this by implementing a Neural Network (NN) controller that is trained only on a family of circular reference trajectories. The sensors used are RTK-GPS and IMU, the latter for providing heading. The NN controller is trained using either a human driver (via human in the loop simulation), or a Model Predictive Control (MPC) strategy. We demonstrate these two approaches in conjunction with two operation scenarios: the vehicle follows a waypoint-defined trajectory at constant speed; and the vehicle follows a speed profile that changes along the vehicle's waypoint-defined trajectory. The primary contribution of this work is the demonstration of Zero-Shot Transfer in conjunction with a novel feed-forward NN controller trained using a general purpose, in-house developed simulation platform.

Understanding complex chemical systems -- such as biomolecules, catalysts, and novel materials -- is a central goal of quantum simulations. Near-term strategies hinge on the use of variational quantum eigensolver (VQE) algorithms combined with a suitable ansatz. However, straightforward application of many chemically-inspired ansatze yields prohibitively deep circuits. In this work, we employ several circuit optimization methods tailored for trapped-ion quantum devices to enhance the feasibility of intricate chemical simulations. The techniques aim to lessen the depth of the unitary coupled cluster with singles and doubles (uCCSD) ansatz's circuit compilation, a considerable challenge on current noisy quantum devices. Furthermore, we use symmetry-inspired classical post-selection methods to further refine the outcomes and minimize errors in energy measurements, without adding quantum overhead. Our strategies encompass optimal mapping from orbital to qubit, term reordering to minimize entangling gates, and the exploitation of molecular spin and point group symmetry to eliminate redundant parameters. The inclusion of error mitigation via post-selection based on known molecular symmetries improves the results to near milli-Hartree accuracy. These methods, when applied to a benzene molecule simulation, enabled the construction of an 8-qubit circuit with 69 two-qubit entangling operations, pushing the limits for variational quantum eigensolver (VQE) circuits executed on quantum hardware to date.

Domain generalization (DG) is about learning models that generalize well to new domains that are related to, but different from, the training domain(s). It is a fundamental problem in machine learning and has attracted much attention in recent years. A large number of approaches have been proposed. Different approaches are motivated from different perspectives, making it difficult to gain an overall understanding of the area. In this paper, we propose a causal framework for domain generalization and present an understanding of common DG approaches in the framework. Our work sheds new lights on the following questions: (1) What are the key ideas behind each DG method? (2) Why is it expected to improve generalization to new domains theoretically? (3) How are different DG methods related to each other and what are relative advantages and limitations? By providing a unified perspective on DG, we hope to help researchers better understand the underlying principles and develop more effective approaches for this critical problem in machine learning.

The advent of large language models (LLMs) has revolutionized natural language processing, enabling the generation of coherent and contextually relevant human-like text. As LLMs increasingly power conversational agents used by the general public world-wide, the synthetic personality embedded in these models, by virtue of training on large amounts of human data, is becoming increasingly important. Since personality is a key factor determining the effectiveness of communication, we present a comprehensive method for administering and validating personality tests on widely-used LLMs, as well as for shaping personality in the generated text of such LLMs. Applying this method, we found: 1) personality measurements in the outputs of some LLMs under specific prompting configurations are reliable and valid; 2) evidence of reliability and validity of synthetic LLM personality is stronger for larger and instruction fine-tuned models; and 3) personality in LLM outputs can be shaped along desired dimensions to mimic specific human personality profiles. We discuss application and ethical implications of the measurement and shaping method, in particular regarding responsible AI.

Machine learning systems produce biased results towards certain demographic groups, known as the fairness problem. Recent approaches to tackle this problem learn a latent code (i.e., representation) through disentangled representation learning and then discard the latent code dimensions correlated with sensitive attributes (e.g., gender). Nevertheless, these approaches may suffer from incomplete disentanglement and overlook proxy attributes (proxies for sensitive attributes) when processing real-world data, especially for unstructured data, causing performance degradation in fairness and loss of useful information for downstream tasks. In this paper, we propose a novel fairness framework that performs debiasing with regard to both sensitive attributes and proxy attributes, which boosts the prediction performance of downstream task models without complete disentanglement. The main idea is to, first, leverage gradient-based explanation to find two model focuses, 1) one focus for predicting sensitive attributes and 2) the other focus for predicting downstream task labels, and second, use them to perturb the latent code that guides the training of downstream task models towards fairness and utility goals. We show empirically that our framework works with both disentangled and non-disentangled representation learning methods and achieves better fairness-accuracy trade-off on unstructured and structured datasets than previous state-of-the-art approaches.

Domain generalization (DG) is about training models that generalize well under domain shift. Previous research on DG has been conducted mostly in single-source or multi-source settings. In this paper, we consider a third, lesser-known setting where a training domain is endowed with a collection of pairs of examples that share the same semantic information. Such semantic sharing (SS) pairs can be created via data augmentation and then utilized for consistency regularization (CR). We present a theory showing CR is conducive to DG and propose a novel CR method called Logit Attribution Matching (LAM). We conduct experiments on five DG benchmarks and four pretrained models with SS pairs created by both generic and targeted data augmentation methods. LAM outperforms representative single/multi-source DG methods and various CR methods that leverage SS pairs. The code and data of this project are available at https://github.com/Gaohan123/LAM, Comment: 19 pages, 12 figures. Accepted by Uncertainty in Artificial Intelligence (UAI) 2024

Modeling a robust control system with a precise GPS-based state estimation capability in simulation can be useful in field navigation applications as it allows for testing and validation in a controlled environment. This testing process would enable navigation systems to be developed and optimized in simulation with direct transferability to real-world scenarios. The multi-physics simulation engine Chrono allows for the creation of scenarios that may be difficult or dangerous to replicate in the field, such as extreme weather or terrain conditions. Autonomy Research Testbed (ART), a specialized robotics algorithm testbed, is operated in conjunction with Chrono to develop an MPC control policy as well as an EKF state estimator. This platform enables users to easily integrate custom algorithms in the autonomy stack. This model is initially developed and used in simulation and then tested on a twin vehicle model in reality, to demonstrate the transferability between simulation and reality (also known as Sim2Real)., Comment: 10 pages,5 figures,submitted to ECCOMAS Thematic Conference on Multibody Dynamics

Variational quantum eigensolvers (VQE) are among the most promising approaches for solving electronic structure problems on near-term quantum computers. A critical challenge for VQE in practice is that one needs to strike a balance between the expressivity of the VQE ansatz versus the number of quantum gates required to implement the ansatz, given the reality of noisy quantum operations on near-term quantum computers. In this work, we consider an orbital-optimized pair-correlated approximation to the unitary coupled cluster with singles and doubles (uCCSD) ansatz and report a highly efficient quantum circuit implementation for trapped-ion architectures. We show that orbital optimization can recover significant additional electron correlation energy without sacrificing efficiency through measurements of low-order reduced density matrices (RDMs). In the dissociation of small molecules, the method gives qualitatively accurate predictions in the strongly-correlated regime when running on noise-free quantum simulators. On IonQ's Harmony and Aria trapped-ion quantum computers, we run end-to-end VQE algorithms with up to 12 qubits and 72 variational parameters - the largest full VQE simulation with a correlated wave function on quantum hardware. We find that even without error mitigation techniques, the predicted relative energies across different molecular geometries are in excellent agreement with noise-free simulators.

We discuss a platform that has both software and hardware components, and whose purpose is to support research into characterizing and mitigating the sim-to-real gap in robotics and vehicle autonomy engineering. The software is operating-system independent and has three main components: a simulation engine called Chrono, which supports high-fidelity vehicle and sensor simulation; an autonomy stack for algorithm design and testing; and a development environment that supports visualization and hardware-in-the-loop experimentation. The accompanying hardware platform is a 1/6th scale vehicle augmented with reconfigurable mountings for computing, sensing, and tracking. Since this vehicle platform has a digital twin within the simulation environment, one can test the same autonomy perception, state estimation, or controls algorithms, as well as the processors they run on, in both simulation and reality. A demonstration is provided to show the utilization of this platform for autonomy research. Future work will concentrate on augmenting ART/ATK with support for a full-sized Chevy Bolt EUV, which will be made available to this group in the immediate future., Comment: 4 pages, Presented at IROS 2022 Workshop on Miniature Robot Platforms for Full Scale Autonomous Vehicle Research. arXiv admin note: substantial text overlap with arXiv:2206.06537

Nonlinear dynamics are ubiquitous in science and engineering applications, but the physics of most complex systems is far from being fully understood. Discovering interpretable governing equations from measurement data can help us understand and predict the behavior of complex dynamic systems. Although extensive work has recently been done in this field, robustly distilling explicit model forms from very sparse data with considerable noise remains intractable. Moreover, quantifying and propagating the uncertainty of the identified system from noisy data is challenging, and relevant literature is still limited. To bridge this gap, we develop a novel Bayesian spline learning framework to identify parsimonious governing equations of nonlinear (spatio)temporal dynamics from sparse, noisy data with quantified uncertainty. The proposed method utilizes spline basis to handle the data scarcity and measurement noise, upon which a group of derivatives can be accurately computed to form a library of candidate model terms. The equation residuals are used to inform the spline learning in a Bayesian manner, where approximate Bayesian uncertainty calibration techniques are employed to approximate posterior distributions of the trainable parameters. To promote the sparsity, an iterative sequential-threshold Bayesian learning approach is developed, using the alternative direction optimization strategy to systematically approximate L0 sparsity constraints. The proposed algorithm is evaluated on multiple nonlinear dynamical systems governed by canonical ordinary and partial differential equations, and the merit/superiority of the proposed method is demonstrated by comparison with state-of-the-art methods., Comment: 28 pages, 11 figures

To study the causes of the 2021 Great Resignation, we use text analysis to investigate the changes in work- and quit-related posts between 2018 and 2021 on Reddit. We find that the Reddit discourse evolution resembles the dynamics of the U.S. quit and layoff rates. Furthermore, when the COVID-19 pandemic started, conversations related to working from home, switching jobs, work-related distress, and mental health increased. We distinguish between general work-related and specific quit-related discourse changes using a difference-in-differences method. Our main finding is that mental health and work-related distress topics disproportionally increased among quit-related posts since the onset of the pandemic, likely contributing to the Great Resignation. Along with better labor market conditions, some relief came beginning-to-mid-2021 when these concerns decreased. Our study validates the use of forums such as Reddit for studying emerging economic phenomena in real time, complementing traditional labor market surveys and administrative data.

In the past decade, with the development of big data technology, an increasing amount of patient information has been stored as electronic health records (EHRs). Leveraging these data, various doctor recommendation systems have been proposed. Typically, such studies process the EHR data in a flat-structured manner, where each encounter was treated as an unordered set of features. Nevertheless, the heterogeneous structured information such as service sequence stored in claims shall not be ignored. This paper presents a doctor recommendation system with time embedding to reconstruct the potential connections between patients and doctors using heterogeneous graph attention network. Besides, to address the privacy issue of patient data sharing crossing hospitals, a federated decentralized learning method based on a minimization optimization model is also proposed. The graph-based recommendation system has been validated on a EHR dataset. Compared to baseline models, the proposed method improves the AUC by up to 6.2%. And our proposed federated-based algorithm not only yields the fictitious fusion center's performance but also enjoys a convergence rate of O(1/T).