Search

Compartmental models, especially the Susceptible-Infected-Removed (SIR) model, have long been used to understand the behaviour of various diseases. Allowing parameters, such as the transmission rate, to be time-dependent functions makes it possible to adjust for and make inferences about changes in the process due to mitigation strategies or evolutionary changes of the infectious agent. In this article, we attempt to build a nonparametric inference framework for stochastic SIR models with time dependent infection rate. The framework includes three main steps: likelihood approximation, parameter estimation and confidence interval construction. The likelihood function of the stochastic SIR model, which is often intractable, can be approximated using methods such as diffusion approximation or tau leaping. The infection rate is modelled by a B-spline basis whose knot location and number of knots are determined by a fast knot placement method followed by a criterion-based model selection procedure. Finally, a point-wise confidence interval is built using a parametric bootstrap procedure. The performance of the framework is observed through various settings for different epidemic patterns. The model is then applied to the Ontario COVID-19 data across multiple waves.

Despite being widely applied due to their exceptional capabilities, Large Language Models (LLMs) have been proven to be vulnerable to backdoor attacks. These attacks introduce targeted vulnerabilities into LLMs by poisoning training samples and full-parameter fine-tuning. However, this kind of backdoor attack is limited since they require significant computational resources, especially as the size of LLMs increases. Besides, parameter-efficient fine-tuning (PEFT) offers an alternative but the restricted parameter updating may impede the alignment of triggers with target labels. In this study, we first verify that backdoor attacks with PEFT may encounter challenges in achieving feasible performance. To address these issues and improve the effectiveness of backdoor attacks with PEFT, we propose a novel backdoor attack algorithm from weak to strong based on contrastive knowledge distillation (W2SAttack). Specifically, we poison small-scale language models through full-parameter fine-tuning to serve as the teacher model. The teacher model then covertly transfers the backdoor to the large-scale student model through contrastive knowledge distillation, which employs PEFT. Theoretical analysis reveals that W2SAttack has the potential to augment the effectiveness of backdoor attacks. We demonstrate the superior performance of W2SAttack on classification tasks across four language models, four backdoor attack algorithms, and two different architectures of teacher models. Experimental results indicate success rates close to 100% for backdoor attacks targeting PEFT.

Text-to-SQL parsing and end-to-end question answering (E2E TQA) are two main approaches for Table-based Question Answering task. Despite success on multiple benchmarks, they have yet to be compared and their synergy remains unexplored. In this paper, we identify different strengths and weaknesses through evaluating state-of-the-art models on benchmark datasets: Text-to-SQL demonstrates superiority in handling questions involving arithmetic operations and long tables; E2E TQA excels in addressing ambiguous questions, non-standard table schema, and complex table contents. To combine both strengths, we propose a Synergistic Table-based Question Answering approach that integrate different models via answer selection, which is agnostic to any model types. Further experiments validate that ensembling models by either feature-based or LLM-based answer selector significantly improves the performance over individual models., Comment: EMNLP 2024

Code completion is essential in software development, helping developers by predicting code snippets based on context. Among completion tasks, Method Body Completion (MBC) is particularly challenging as it involves generating complete method bodies based on their signatures and context. This task becomes significantly harder in large repositories, where method bodies must integrate repositoryspecific elements such as custom APIs, inter-module dependencies, and project-specific conventions. In this paper, we introduce RAMBO, a novel RAG-based approach for repository-level MBC. Instead of retrieving similar method bodies, RAMBO identifies essential repository-specific elements, such as classes, methods, and variables/fields, and their relevant usages. By incorporating these elements and their relevant usages into the code generation process, RAMBO ensures more accurate and contextually relevant method bodies. Our experimental results with leading code LLMs across 40 Java projects show that RAMBO significantly outperformed the state-of-the-art repository-level MBC approaches, with the improvements of up to 46% in BLEU, 57% in CodeBLEU, 36% in Compilation Rate, and up to 3X in Exact Match. Notably, RAMBO surpassed RepoCoder Oracle method by up to 12% in Exact Match, setting a new benchmark for repository-level MBC.

Large Language Models (LLMs) have demonstrated remarkable performance through supervised fine-tuning or in-context learning using gold labels. However, this paradigm is limited by the availability of gold labels, while in certain scenarios, LLMs may need to perform tasks that are too complex for humans to provide such labels. To tackle this challenge, this study explores whether solely utilizing unlabeled data can elicit strong model capabilities. We propose a new paradigm termed zero-to-strong generalization. We iteratively prompt LLMs to annotate unlabeled data and retain high-quality labels by filtering. Surprisingly, we obverse that this iterative process gradually unlocks LLMs' potential on downstream tasks. Our experiments on extensive classification and reasoning tasks confirm the effectiveness of our proposed framework. Our analysis indicates that this paradigm is effective for both in-context learning and fine-tuning, and for various model sizes., Comment: 15 pages

Autonomous driving systems (ADSs) have undergone remarkable development and are increasingly employed in safety-critical applications. However, recently reported data on fatal accidents involving ADSs suggests that the desired level of safety has not yet been fully achieved. Consequently, there is a growing need for more comprehensive and targeted testing approaches to ensure safe driving. Scenarios from real-world accident reports provide valuable resources for ADS testing, including critical scenarios and high-quality seeds. However, existing scenario reconstruction methods from accident reports often exhibit limited accuracy in information extraction. Moreover, due to the diversity and complexity of road environments, matching current accident information with the simulation map data for reconstruction poses significant challenges. In this paper, we design and implement SoVAR, a tool for automatically generating road-generalizable scenarios from accident reports. SoVAR utilizes well-designed prompts with linguistic patterns to guide the large language model in extracting accident information from textual data. Subsequently, it formulates and solves accident-related constraints in conjunction with the extracted accident information to generate accident trajectories. Finally, SoVAR reconstructs accident scenarios on various map structures and converts them into test scenarios to evaluate its capability to detect defects in industrial ADSs. We experiment with SoVAR, using accident reports from the National Highway Traffic Safety Administration's database to generate test scenarios for the industrial-grade ADS Apollo. The experimental findings demonstrate that SoVAR can effectively generate generalized accident scenarios across different road structures. Furthermore, the results confirm that SoVAR identified 5 distinct safety violation types that contributed to the crash of Baidu Apollo.

Temperate exoplanets between the sizes of Earth and Neptune, known as "sub-Neptunes", have emerged as intriguing targets for astrobiology. It is unknown whether these planets resemble Earth-like terrestrial worlds with a habitable surface, Neptune-like giant planets with deep atmospheres and no habitable surface, or something exotic in between. Recent JWST transmission spectroscopy observations of the canonical sub-Neptune K2-18 b revealed ~1% CH4, ~1% CO2, and a non-detection of CO in the atmosphere. While previous studies have proposed that the observed atmospheric composition could help constrain the lower atmosphere conditions and determine the interior structure of sub-Neptunes like K2-18 b, the possible interactions between the atmosphere and a hot, supercritical water ocean at its base remain unexplored. In this work, we investigate whether a global supercritical water ocean, resembling a planetary-scale hydrothermal system, can explain these observations on K2-18 b-like sub-Neptunes through equilibrium aqueous geochemical calculations. We find that the observed atmospheric CH4/CO2 ratio implies a minimum ocean temperature of ~715 K, whereas the corresponding CO/CO2 ratio allows ocean temperatures up to ~1060 K. These results indicate that a global supercritical water ocean on K2-18 b is plausible. While life cannot survive in this ocean, this work represents the first step towards understanding how a global supercritical water ocean may influence observable atmospheric characteristics on volatile-rich sub-Neptunes. Future observations with better constrained NH3 and CO mixing ratios could further help distinguish between possible interior compositions of K2-18 b., Comment: 15 pages, 5 figures, 1 table

Autism Spectrum Disorder (ASD) can affect individuals at varying degrees of intensity, from challenges in overall health, communication, and sensory processing, and this often begins at a young age. Thus, it is critical for medical professionals to be able to accurately diagnose ASD in young children, but doing so is difficult. Deep learning can be responsibly leveraged to improve productivity in addressing this task. The availability of data, however, remains a considerable obstacle. Hence, in this work, we introduce the Video ASD dataset--a dataset that contains video frame convolutional and attention map feature data--to foster further progress in the task of ASD classification. The original videos showcase children reacting to chemo-sensory stimuli, among auditory, touch, and vision This dataset contains the features of the frames spanning 2,467 videos, for a total of approximately 1.4 million frames. Additionally, head pose angles are included to account for head movement noise, as well as full-sentence text labels for the taste and smell videos that describe how the facial expression changes before, immediately after, and long after interaction with the stimuli. In addition to providing features, we also test foundation models on this data to showcase how movement noise affects performance and the need for more data and more complex labels.

The advancement of Large Language Models (LLMs) for domain applications in fields such as materials science and engineering depends on the development of fine-tuning strategies that adapt models for specialized, technical capabilities. In this work, we explore the effects of Continued Pretraining (CPT), Supervised Fine-Tuning (SFT), and various preference-based optimization approaches, including Direct Preference Optimization (DPO) and Odds Ratio Preference Optimization (ORPO), on fine-tuned LLM performance. Our analysis shows how these strategies influence model outcomes and reveals that the merging of multiple fine-tuned models can lead to the emergence of capabilities that surpass the individual contributions of the parent models. We find that model merging leads to new functionalities that neither parent model could achieve alone, leading to improved performance in domain-specific assessments. Experiments with different model architectures are presented, including Llama 3.1 8B and Mistral 7B models, where similar behaviors are observed. Exploring whether the results hold also for much smaller models, we use a tiny LLM with 1.7 billion parameters and show that very small LLMs do not necessarily feature emergent capabilities under model merging, suggesting that model scaling may be a key component. In open-ended yet consistent chat conversations between a human and AI models, our assessment reveals detailed insights into how different model variants perform and show that the smallest model achieves a high intelligence score across key criteria including reasoning depth, creativity, clarity, and quantitative precision. Other experiments include the development of image generation prompts based on disparate biological material design concepts, to create new microstructures, architectural concepts, and urban design based on biological materials-inspired construction principles.

Large language models (LLMs) face significant challenges in handling long-context tasks because of their limited effective context window size during pretraining, which restricts their ability to generalize over extended sequences. Meanwhile, extending the context window in LLMs through post-pretraining is highly resource-intensive. To address this, we introduce LongRecipe, an efficient training strategy for extending the context window of LLMs, including impactful token analysis, position index transformation, and training optimization strategies. It simulates long-sequence inputs while maintaining training efficiency and significantly improves the model's understanding of long-range dependencies. Experiments on three types of LLMs show that LongRecipe can utilize long sequences while requiring only 30% of the target context window size, and reduces computational training resource over 85% compared to full sequence training. Furthermore, LongRecipe also preserves the original LLM's capabilities in general tasks. Ultimately, we can extend the effective context window of open-source LLMs from 8k to 128k, achieving performance close to GPT-4 with just one day of dedicated training using a single GPU with 80G memory. Our code is released at https://github.com/zhiyuanhubj/LongRecipe., Comment: Work in Progress

Indoor photovoltaics (IPVs) provide an increasingly promising solution for powering Internet-of-Things smart devices, which has led to a surge in IPV research and the development of new IPV technologies. However, the diverse lighting scenarios adopted in IPV research pose unique challenges in characterization, reporting, and benchmarking, which may obscure genuine performance improvements and result in inaccurate conclusions due to characterization errors. Spectral variations among artificial light sources further complicate benchmarking. This study provides a comprehensive, quantitative analysis of these challenges, investigating them through experimental characterization of IPVs covering a broad performance parameter space, including c-Si, a-Si:H, perovskite, and organic devices. We reveal that many of these challenges can lead to unacceptable error levels. A particularly critical issue is the angular interplay among the light source, measuring device, and IPV device, which compromises accuracy under diffuse illumination. To address these challenges, we evaluate practical protocols to overcome angular issues and enable benchmarking against standardized spectral conditions. To facilitate the implementation of our findings, we provide comprehensive checklists for accurate IPV characterization, reporting, and benchmarking. We anticipate that our analyses and guidelines will stimulate further advancements in IPV technologies by ensuring reliable performance evaluation, thereby facilitating the realization of IPVs' full potential.

Open-Vocab 3D Instance Segmentation methods (OV-3DIS) have recently demonstrated their ability to generalize to unseen objects. However, these methods still depend on predefined class names during testing, restricting the autonomy of agents. To mitigate this constraint, we propose a novel problem termed Open-Ended 3D Instance Segmentation (OE-3DIS), which eliminates the necessity for predefined class names during testing. Moreover, we contribute a comprehensive set of strong baselines, derived from OV-3DIS approaches and leveraging 2D Multimodal Large Language Models. To assess the performance of our OE-3DIS system, we introduce a novel Open-Ended score, evaluating both the semantic and geometric quality of predicted masks and their associated class names, alongside the standard AP score. Our approach demonstrates significant performance improvements over the baselines on the ScanNet200 and ScanNet++ datasets. Remarkably, our method surpasses the performance of Open3DIS, the current state-of-the-art method in OV-3DIS, even in the absence of ground-truth object class names.

SPIDER is a balloon-borne instrument designed to map the cosmic microwave background at degree-angular scales in the presence of Galactic foregrounds. SPIDER has mapped a large sky area in the Southern Hemisphere using more than 2000 transition-edge sensors (TESs) during two NASA Long Duration Balloon flights above the Antarctic continent. During its first flight in January 2015, SPIDER observed in the 95 GHz and 150 GHz frequency bands, setting constraints on the B-mode signature of primordial gravitational waves. Its second flight in the 2022-23 season added new receivers at 280 GHz, each using an array of TESs coupled to the sky through feedhorns formed from stacks of silicon wafers. These receivers are optimized to produce deep maps of polarized Galactic dust emission over a large sky area, providing a unique data set with lasting value to the field. In this work, we describe the instrument's performance during SPIDER's second flight., Comment: Submitted to SPIE Astronomical Telescopes + Instrumentation 2024, JATIS

Quantum computing has emerged as a powerful tool for solving complex problems intractable for classical computers, particularly in popular fields such as cryptography, optimization, and neurocomputing. In this paper, we present a new quantum-based approach named the Hierarchical Quantum Control Gates (HQCG) method for efficient understanding of Functional Magnetic Resonance Imaging (fMRI) data. This approach includes two novel modules: the Local Quantum Control Gate (LQCG) and the Global Quantum Control Gate (GQCG), which are designed to extract local and global features of fMRI signals, respectively. Our method operates end-to-end on a quantum machine, leveraging quantum mechanics to learn patterns within extremely high-dimensional fMRI signals, such as 30,000 samples which is a challenge for classical computers. Empirical results demonstrate that our approach significantly outperforms classical methods. Additionally, we found that the proposed quantum model is more stable and less prone to overfitting than the classical methods., Comment: Accepted to IEEE Workshop on Signal Processing Systems (SiPS 2024)

SuperBIT was a 0.5-meter near-ultraviolet to near-infrared wide-field telescope that launched on a NASA superpressure balloon into the stratosphere from New Zealand for a 45-night flight. SuperBIT acquired multi-band images of galaxy clusters to study the properties of dark matter using weak gravitational lensing. We provide an overview of the instrument and its various subsystems. We then present the instrument performance from the flight, including the telescope and image stabilization system, the optical system, the power system, and the thermal system. SuperBIT successfully met the instrument's technical requirements, achieving a telescope pointing stability of 0.34 +/- 0.10 arcseconds, a focal plane image stability of 0.055 +/- 0.027 arcseconds, and a PSF FWHM of ~ 0.35 arcseconds over 5-minute exposures throughout the 45-night flight. The telescope achieved a near-diffraction limited point-spread function in all three science bands (u, b, and g). SuperBIT served as a pathfinder to the GigaBIT observatory, which will be a 1.34-meter near-ultraviolet to near-infrared balloon-borne telescope., Comment: 17 pages, 25 pages, published in the Astronomical Journal

Fuzzy dark matter (FDM) is a compelling candidate for dark matter, offering a natural explanation for the structure of diffuse low-mass haloes. However, the canonical FDM model with a mass of $10^{-22}~{\rm eV}$ encounters challenges in reproducing the observed diversity of dwarf galaxies, except for possibly scenarios where strong galactic feedback is invoked. The introduction of multiple-field FDM can provide a potential resolution to this diversity issue. The theoretical plausibility of this dark matter model is also enhanced by the fact that multiple axion species with logarithmically-distributed mass spectrum exist as a generic prediction of string theory. In this paper we consider the axiverse hypothesis and investigate non-linear structure formation in the two-field fuzzy dark matter (2FDM) model. Our cosmological simulation with an unprecedented resolution and self-consistent initial conditions reveals the diverse structures of dark matter haloes in the 2FDM model for the first time. Depending on the formation time and local tidal activities, late-time haloes can host solitons of nested cores or solitons of one dominant species., Comment: 9 pages, 5 figures. Comments welcome!

Quantum state tomography is a crucial technique for characterizing the state of a quantum system, which is essential for many applications in quantum technologies. In recent years, there has been growing interest in leveraging neural networks to enhance the efficiency and accuracy of quantum state tomography. Still, many of them did not include mixed quantum state, since pure states are arguably less common in practical situations. In this research paper, we present two neural networks based approach for both pure and mixed quantum state tomography: Restricted Feature Based Neural Network and Mixed States Conditional Generative Adversarial Network, evaluate its effectiveness in comparison to existing neural based methods. We demonstrate that our proposed methods can achieve state-of-the-art results in reconstructing mixed quantum states from experimental data. Our work highlights the potential of neural networks in revolutionizing quantum state tomography and facilitating the development of quantum technologies., Comment: 10 pages, 5 figures, 17 illustration, 1 table

Reinforcement Learning (RL) agents demonstrating proficiency in a training environment exhibit vulnerability to adversarial perturbations in input observations during deployment. This underscores the importance of building a robust agent before its real-world deployment. To alleviate the challenging point, prior works focus on developing robust training-based procedures, encompassing efforts to fortify the deep neural network component's robustness or subject the agent to adversarial training against potent attacks. In this work, we propose a novel method referred to as Transformed Input-robust RL (TIRL), which explores another avenue to mitigate the impact of adversaries by employing input transformation-based defenses. Specifically, we introduce two principles for applying transformation-based defenses in learning robust RL agents: (1) autoencoder-styled denoising to reconstruct the original state and (2) bounded transformations (bit-depth reduction and vector quantization (VQ)) to achieve close transformed inputs. The transformations are applied to the state before feeding it into the policy network. Extensive experiments on multiple MuJoCo environments demonstrate that input transformation-based defenses, i.e., VQ, defend against several adversaries in the state observations. The official code is available at https://github.com/tunglm2203/tirl, Comment: 12 pages (Code: https://github.com/tunglm2203/tirl)

Magnetic particle imaging (MPI) is an emerging imaging technique with many applications and a very active field of research. This app provides users with the opportunity to develop some intuition about the inner workings of MPI as it is being researched through NIST's Thermal MagIC project in an interactive and fun way. Users can vary different experimental and post-processing parameters to see how the image quality and particle reconstruction changes for different measurement conditions., Comment: 7 pages, 6 figures

In online learning, the ability to provide quick and accurate feedback to learners is crucial. In skill-based learning, learners need to understand the underlying concepts and mechanisms of a skill to be able to apply it effectively. While videos are a common tool in online learning, they cannot comprehend or assess the skills being taught. Additionally, while Generative AI methods are effective in searching and retrieving answers from a text corpus, it remains unclear whether these methods exhibit any true understanding. This limits their ability to provide explanations of skills or help with problem-solving. This paper proposes a novel approach that merges Cognitive AI and Generative AI to address these challenges. We employ a structured knowledge representation, the TMK (Task-Method-Knowledge) model, to encode skills taught in an online Knowledge-based AI course. Leveraging techniques such as Large Language Models, Chain-of-Thought, and Iterative Refinement, we outline a framework for generating reasoned explanations in response to learners' questions about skills., Comment: 9 pages, 6 figures, 1 table

Continuing advances in Large Language Models (LLMs) in artificial intelligence offer important capacities in intuitively accessing and using medical knowledge in many contexts, including education and training as well as assessment and treatment. Most of the initial literature on LLMs in medicine has emphasized that LLMs are unsuitable for medical use because they are non-deterministic, may provide incorrect or harmful responses, and cannot be regulated to assure quality control. If these issues could be corrected, optimizing LLM technology could benefit patients and physicians by providing affordable, point-of-care medical knowledge. Our proposed framework refines LLM responses by restricting their primary knowledge base to domain-specific datasets containing validated medical information. Additionally, we introduce an actor-critic LLM prompting protocol based on active inference principles of human cognition, where a Therapist agent initially responds to patient queries, and a Supervisor agent evaluates and adjusts responses to ensure accuracy and reliability. We conducted a validation study where expert cognitive behaviour therapy for insomnia (CBT-I) therapists evaluated responses from the LLM in a blind format. Experienced human CBT-I therapists assessed responses to 100 patient queries, comparing LLM-generated responses with appropriate and inappropriate responses crafted by experienced CBT-I therapists. Results showed that LLM responses received high ratings from the CBT-I therapists, often exceeding those of therapist-generated appropriate responses. This structured approach aims to integrate advanced LLM technology into medical applications, meeting regulatory requirements for establishing the safe and effective use of special purpose validated LLMs in medicine., Comment: 25 pages, 4 figures

We compare the abundant prolate shaped galaxies reported beyond z$>$3 in deep JWST surveys, with the predicted {\it stellar} appearance of young galaxies in detailed hydro-simulations of three main dark matter contenders: Cold (CDM), Wave/Fuzzy ($\psi$DM) and Warm Dark Matter (WDM). We find the observed galaxy images closely resemble the elongated stellar appearance of young galaxies predicted for both $\psi$DM and WDM, during the first $\simeq$ 500Myr while material steadily accretes from long, smooth filaments. The dark mater halos of WDM and $\psi$DM also have pronounced, prolate elongation similar to the stars, indicating a shared, highly triaxial equilibrium. This is unlike CDM where the early stellar morphology is mainly spheroidal formed from fragmented filaments with frequent merging, resulting in modest triaxiality. Quantitatively, the excess of prolate galaxies observed by JWST matches well WDM and $\psi$DM for particle masses of 1.4KeV and $2.5\times 10^{-22}$ eV respectively. For CDM, several visible subhalos are typically predicted to orbit within the virial radius of each galaxy from subhalo accretion, whereas merging is initially rare for WDM and $\psi$DM. We also verify with our simulations that $\psi$DM may be distinguished from WDM by the form of the core, which is predicted to be smooth and centered for WDM, but is a dense soliton for $\psi$DM traced by stars and measurably offset from the galaxy center by random wave perturbations in the simulations. We emphasise that long smooth filaments absent of galaxies may prove detectable with JWST, traced by stars and gas with comoving lengths of 150kpc predicted at z$\simeq$10, depending on the particle mass of $\psi$DM or WDM., Comment: 20 pages, 14 figures

This paper examines the pricing issue of margin-call stock loans with finite maturities under the Black-Scholes-Merton framework. In particular, using a Fourier Sine transform method, we reduce the partial differential equation governing the price of a margin-call stock loan into an ordinary differential equation, the solution of which can be easily found (in the Fourier Sine space) and analytically inverted into the original space. As a result, we obtain an integral representation of the value of the stock loan in terms of the unknown optimal exit prices, which are, in turn, governed by a Volterra integral equation. We thus can break the pricing problem of margin-call stock loans into two steps: 1) finding the optimal exit prices by solving numerically the governing Volterra integral equation and 2) calculating the values of margin-call stock loans based on the obtained optimal exit prices. By validating and comparing with other available numerical methods, we show that our proposed numerical scheme offers a reliable and efficient way to calculate the service fee of a margin-call stock loan contract, track the contract value over time, and compute the level of stock price above which it is optimal to exit the contract. The effects of the margin-call feature on the loan contract are also examined and quantified.

The Super-pressure Balloon-borne Imaging Telescope (SuperBIT) is a near-diffraction-limited 0.5m telescope that launched via NASA's super-pressure balloon technology on April 16, 2023. SuperBIT achieved precise pointing control through the use of three nested frames in conjunction with an optical Fine Guidance System (FGS), resulting in an average image stability of 0.055" over 300-second exposures. The SuperBIT FGS includes a tip-tilt fast-steering mirror that corrects for jitter on a pair of focal plane star cameras. In this paper, we leverage the empirical data from SuperBIT's successful 45-night stratospheric mission to inform the FGS design for the next-generation balloon-borne telescope. The Gigapixel Balloon-borne Imaging Telescope (GigaBIT) is designed to be a 1.35m wide-field, high resolution imaging telescope, with specifications to extend the scale and capabilities beyond those of its predecessor SuperBIT. A description and analysis of the SuperBIT FGS will be presented along with methodologies for extrapolating this data to enhance GigaBIT's FGS design and fine pointing control algorithm. We employ a systems engineering approach to outline and formalize the design constraints and specifications for GigaBIT's FGS. GigaBIT, building on the SuperBIT legacy, is set to enhance high-resolution astronomical imaging, marking a significant advancement in the field of balloon-borne telescopes., Comment: 13 pages, 7 figures, SPIE Astronomical Telescopes + Instrumentation 2024

We apply constant imaginary offsets to the path integral for a reduction of the sign problem in the Hubbard model. These simple transformations enhance the quality of results from HMC calculations without compromising the speed of the algorithm. This method enables us to efficiently calculate systems that are otherwise inaccessible due to a severe sign problem. To support this claim, we present observables of the C20 and C60 fullerenes. Furthermore, we demonstrate that at a certain offset, the sign problem is completely lifted in the limit of large chemical potential., Comment: Proceedings for the 40th Lattice conference (2023)

Businesses need to query visually rich documents (VRDs) like receipts, medical records, and insurance forms to make decisions. Existing techniques for extracting entities from VRDs struggle with new layouts or require extensive pre-training data. We introduce VRDSynth, a program synthesis method to automatically extract entity relations from multilingual VRDs without pre-training data. To capture the complexity of VRD domain, we design a domain-specific language (DSL) to capture spatial and textual relations to describe the synthesized programs. Along with this, we also derive a new synthesis algorithm utilizing frequent spatial relations, search space pruning, and a combination of positive, negative, and exclusive programs to improve coverage. We evaluate VRDSynth on the FUNSD and XFUND benchmarks for semantic entity linking, consisting of 1,592 forms in 8 languages. VRDSynth outperforms state-of-the-art pre-trained models (LayoutXLM, InfoXLMBase, and XLMRobertaBase) in 5, 6, and 7 out of 8 languages, respectively, improving the F1 score by 42% over LayoutXLM in English. To test the extensibility of the model, we further improve VRDSynth with automated table recognition, creating VRDSynth(Table), and compare it with extended versions of the pre-trained models, InfoXLM(Large) and XLMRoberta(Large). VRDSynth(Table) outperforms these baselines in 4 out of 8 languages and in average F1 score. VRDSynth also significantly reduces memory footprint (1M and 380MB vs. 1.48GB and 3GB for LayoutXLM) while maintaining similar time efficiency., Comment: Accepted in ISSTA'24

Data quality stands at the forefront of deciding the effectiveness of video-language representation learning. However, video-text pairs in previous data typically do not align perfectly with each other, which might lead to video-language representations that do not accurately reflect cross-modal semantics. Moreover, previous data also possess an uneven distribution of concepts, thereby hampering the downstream performance across unpopular subjects. To address these problems, we propose a contrastive objective with a subtractive angular margin to regularize cross-modal representations in their effort to reach perfect similarity. Furthermore, to adapt to the non-uniform concept distribution, we propose a multi-layer perceptron (MLP)-parameterized weighting function that maps loss values to sample weights which enable dynamic adjustment of the model's focus throughout the training. With the training guided by a small amount of unbiased meta-data and augmented by video-text data generated by large vision-language model, we improve video-language representations and achieve superior performances on commonly used video question answering and text-video retrieval datasets., Comment: Accepted to ECCV 2024

While diffusion models excel at conditional generating high-quality images, prior works in discrete diffusion models were not evaluated on conditional long-text generation. In this work, we address the limitations of prior discrete diffusion models for conditional long-text generation, particularly in long sequence-to-sequence tasks such as abstractive summarization. Despite fast decoding speeds compared to autoregressive methods, previous diffusion models failed on the abstractive summarization task due to the incompatibility between the backbone architectures and the random noising process. To overcome these challenges, we introduce a novel semantic-aware noising process that enables Transformer backbones to handle long sequences effectively. Additionally, we propose CrossMamba, an adaptation of the Mamba model to the encoder-decoder paradigm, which integrates seamlessly with the random absorbing noising process. Our approaches achieve state-of-the-art performance on three benchmark summarization datasets: Gigaword, CNN/DailyMail, and Arxiv, outperforming existing discrete diffusion models on ROUGE metrics as well as possessing much faster speed in inference compared to autoregressive models.

The development of Natural Language Processing (NLI) datasets and models has been inspired by innovations in annotation design. With the rapid development of machine learning models today, the performance of existing machine learning models has quickly reached state-of-the-art results on a variety of tasks related to natural language processing, including natural language inference tasks. By using a pre-trained model during the annotation process, it is possible to challenge current NLI models by having humans produce premise-hypothesis combinations that the machine model cannot correctly predict. To remain attractive and challenging in the research of natural language inference for Vietnamese, in this paper, we introduce the adversarial NLI dataset to the NLP research community with the name ViANLI. This data set contains more than 10K premise-hypothesis pairs and is built by a continuously adjusting process to obtain the most out of the patterns generated by the annotators. ViANLI dataset has brought many difficulties to many current SOTA models when the accuracy of the most powerful model on the test set only reached 48.4%. Additionally, the experimental results show that the models trained on our dataset have significantly improved the results on other Vietnamese NLI datasets.

In this paper, we introduce UniBridge (Cross-Lingual Transfer Learning with Optimized Embeddings and Vocabulary), a comprehensive approach developed to improve the effectiveness of Cross-Lingual Transfer Learning, particularly in languages with limited resources. Our approach tackles two essential elements of a language model: the initialization of embeddings and the optimal vocabulary size. Specifically, we propose a novel embedding initialization method that leverages both lexical and semantic alignment for a language. In addition, we present a method for systematically searching for the optimal vocabulary size, ensuring a balance between model complexity and linguistic coverage. Our experiments across multilingual datasets show that our approach greatly improves the F1-Score in several languages. UniBridge is a robust and adaptable solution for cross-lingual systems in various languages, highlighting the significance of initializing embeddings and choosing the right vocabulary size in cross-lingual environments., Comment: First two authors contribute equally. Accepted at ACL 2024

We present a Hamiltonian Monte Carlo study of doped perylene $\mathrm{C}_{20}\mathrm{H}_{12}$ described with the Hubbard model. Doped perylene can be used for organic light-emitting diodes (OLEDs) or as acceptor material in organic solar cells. Therefore, central to this study is a scan over charge chemical potential. A variational basis of operators allows for the extraction of the single-particle spectrum through a mostly automatic fitting procedure. Finite chemical potential simulations suffer from a sign problem which we ameliorate through contour deformation. The on-site interaction is kept at $U/\kappa = 2$. Discretization effects are handled through a continuum limit extrapolation. Our first-principles calculation shows significant deviation from non-interacting results especially at large chemical potentials.

Large Language Models (LLMs), which bridge the gap between human language understanding and complex problem-solving, achieve state-of-the-art performance on several NLP tasks, particularly in few-shot and zero-shot settings. Despite the demonstrable efficacy of LLMs, due to constraints on computational resources, users have to engage with open-source language models or outsource the entire training process to third-party platforms. However, research has demonstrated that language models are susceptible to potential security vulnerabilities, particularly in backdoor attacks. Backdoor attacks are designed to introduce targeted vulnerabilities into language models by poisoning training samples or model weights, allowing attackers to manipulate model responses through malicious triggers. While existing surveys on backdoor attacks provide a comprehensive overview, they lack an in-depth examination of backdoor attacks specifically targeting LLMs. To bridge this gap and grasp the latest trends in the field, this paper presents a novel perspective on backdoor attacks for LLMs by focusing on fine-tuning methods. Specifically, we systematically classify backdoor attacks into three categories: full-parameter fine-tuning, parameter-efficient fine-tuning, and no fine-tuning Based on insights from a substantial review, we also discuss crucial issues for future research on backdoor attacks, such as further exploring attack algorithms that do not require fine-tuning, or developing more covert attack algorithms.

Humans use multiple senses to comprehend the environment. Vision and language are two of the most vital senses since they allow us to easily communicate our thoughts and perceive the world around us. There has been a lot of interest in creating video-language understanding systems with human-like senses since a video-language pair can mimic both our linguistic medium and visual environment with temporal dynamics. In this survey, we review the key tasks of these systems and highlight the associated challenges. Based on the challenges, we summarize their methods from model architecture, model training, and data perspectives. We also conduct performance comparison among the methods, and discuss promising directions for future research., Comment: Accepted at ACL 2024 (Findings)

The Vision-Language Foundation Model has recently shown outstanding performance in various perception learning tasks. The outstanding performance of the vision-language model mainly relies on large-scale pre-training datasets and different data augmentation techniques. However, the domain generalization problem of the vision-language foundation model needs to be addressed. This problem has limited the generalizability of the vision-language foundation model to unknown data distributions. In this paper, we introduce a new simple but efficient Diffusion Sampling approach to Domain Generalization (ED-SAM) to improve the generalizability of the vision-language foundation model. Our theoretical analysis in this work reveals the critical role and relation of the diffusion model to domain generalization in the vision-language foundation model. Then, based on the insightful analysis, we introduce a new simple yet effective Transport Transformation to diffusion sampling method. It can effectively generate adversarial samples to improve the generalizability of the foundation model against unknown data distributions. The experimental results on different scales of vision-language pre-training datasets, including CC3M, CC12M, and LAION400M, have consistently shown State-of-the-Art performance and scalability of the proposed ED-SAM approach compared to the other recent methods.

Unsupervised Domain Adaptation has been an efficient approach to transferring the semantic segmentation model across data distributions. Meanwhile, the recent Open-vocabulary Semantic Scene understanding based on large-scale vision language models is effective in open-set settings because it can learn diverse concepts and categories. However, these prior methods fail to generalize across different camera views due to the lack of cross-view geometric modeling. At present, there are limited studies analyzing cross-view learning. To address this problem, we introduce a novel Unsupervised Cross-view Adaptation Learning approach to modeling the geometric structural change across views in Semantic Scene Understanding. First, we introduce a novel Cross-view Geometric Constraint on Unpaired Data to model structural changes in images and segmentation masks across cameras. Second, we present a new Geodesic Flow-based Correlation Metric to efficiently measure the geometric structural changes across camera views. Third, we introduce a novel view-condition prompting mechanism to enhance the view-information modeling of the open-vocabulary segmentation network in cross-view adaptation learning. The experiments on different cross-view adaptation benchmarks have shown the effectiveness of our approach in cross-view modeling, demonstrating that we achieve State-of-the-Art (SOTA) performance compared to prior unsupervised domain adaptation and open-vocabulary semantic segmentation methods.

Video scene graph generation (VidSGG) has emerged as a transformative approach to capturing and interpreting the intricate relationships among objects and their temporal dynamics in video sequences. In this paper, we introduce the new AeroEye dataset that focuses on multi-object relationship modeling in aerial videos. Our AeroEye dataset features various drone scenes and includes a visually comprehensive and precise collection of predicates that capture the intricate relationships and spatial arrangements among objects. To this end, we propose the novel Cyclic Graph Transformer (CYCLO) approach that allows the model to capture both direct and long-range temporal dependencies by continuously updating the history of interactions in a circular manner. The proposed approach also allows one to handle sequences with inherent cyclical patterns and process object relationships in the correct sequential order. Therefore, it can effectively capture periodic and overlapping relationships while minimizing information loss. The extensive experiments on the AeroEye dataset demonstrate the effectiveness of the proposed CYCLO model, demonstrating its potential to perform scene understanding on drone videos. Finally, the CYCLO method consistently achieves State-of-the-Art (SOTA) results on two in-the-wild scene graph generation benchmarks, i.e., PVSG and ASPIRe.

Parameterized Quantum Circuits (PQCs) have been acknowledged as a leading strategy to utilize near-term quantum advantages in multiple problems, including machine learning and combinatorial optimization. When applied to specific tasks, the parameters in the quantum circuits are trained to minimize the target function. Although there have been comprehensive studies to improve the performance of the PQCs on practical tasks, the errors caused by the quantum noise downgrade the performance when running on real quantum computers. In particular, when the quantum state is transformed through multiple quantum circuit layers, the effect of the quantum noise happens cumulatively and becomes closer to the maximally mixed state or complete noise. This paper studies the relationship between the quantum noise and the diffusion model. Then, we propose a novel diffusion-inspired learning approach to mitigate the quantum noise in the PQCs and reduce the error for specific tasks. Through our experiments, we illustrate the efficiency of the learning strategy and achieve state-of-the-art performance on classification tasks in the quantum noise scenarios.

We study a class of optimization problems in the Wasserstein space (the space of probability measures) where the objective function is \emph{nonconvex} along generalized geodesics. When the regularization term is the negative entropy, the optimization problem becomes a sampling problem where it minimizes the Kullback-Leibler divergence between a probability measure (optimization variable) and a target probability measure whose logarithmic probability density is a nonconvex function. We derive multiple convergence insights for a novel {\em semi Forward-Backward Euler scheme} under several nonconvex (and possibly nonsmooth) regimes. Notably, the semi Forward-Backward Euler is just a slight modification of the Forward-Backward Euler whose convergence is -- to our knowledge -- still unknown in our very general non-geodesically-convex setting.

This paper proposes a fully explainable approach to speaker verification (SV), a task that fundamentally relies on individual speaker characteristics. The opaque use of speaker attributes in current SV systems raises concerns of trust. Addressing this, we propose an attribute-based explainable SV system that identifies speakers by comparing personal attributes such as gender, nationality, and age extracted automatically from voice recordings. We believe this approach better aligns with human reasoning, making it more understandable than traditional methods. Evaluated on the Voxceleb1 test set, the best performance of our system is comparable with the ground truth established when using all correct attributes, proving its efficacy. Whilst our approach sacrifices some performance compared to non-explainable methods, we believe that it moves us closer to the goal of transparent, interpretable AI and lays the groundwork for future enhancements through attribute expansion.

Although quantum machine learning has been introduced for a while, its applications in computer vision are still limited. This paper, therefore, revisits the quantum visual encoding strategies, the initial step in quantum machine learning. Investigating the root cause, we uncover that the existing quantum encoding design fails to ensure information preservation of the visual features after the encoding process, thus complicating the learning process of the quantum machine learning models. In particular, the problem, termed "Quantum Information Gap" (QIG), leads to a gap of information between classical and corresponding quantum features. We provide theoretical proof and practical demonstrations of that found and underscore the significance of QIG, as it directly impacts the performance of quantum machine learning algorithms. To tackle this challenge, we introduce a simple but efficient new loss function named Quantum Information Preserving (QIP) to minimize this gap, resulting in enhanced performance of quantum machine learning algorithms. Extensive experiments validate the effectiveness of our approach, showcasing superior performance compared to current methodologies and consistently achieving state-of-the-art results in quantum modeling., Comment: Accepted to Quantum Machine Intelligence

Seeking answers effectively for long videos is essential to build video question answering (videoQA) systems. Previous methods adaptively select frames and regions from long videos to save computations. However, this fails to reason over the whole sequence of video, leading to sub-optimal performance. To address this problem, we introduce a state space layer (SSL) into multi-modal Transformer to efficiently integrate global semantics of the video, which mitigates the video information loss caused by frame and region selection modules. Our SSL includes a gating unit to enable controllability over the flow of global semantics into visual representations. To further enhance the controllability, we introduce a cross-modal compositional congruence (C^3) objective to encourage global semantics aligned with the question. To rigorously evaluate long-form videoQA capacity, we construct two new benchmarks Ego-QA and MAD-QA featuring videos of considerably long length, i.e. 17.5 minutes and 1.9 hours, respectively. Extensive experiments demonstrate the superiority of our framework on these new as well as existing datasets., Comment: Accepted to the main EMNLP 2024 conference

Unsupervised vision clustering, a cornerstone in computer vision, has been studied for decades, yielding significant outcomes across numerous vision tasks. However, these algorithms involve substantial computational demands when confronted with vast amounts of unlabeled data. Conversely, quantum computing holds promise in expediting unsupervised algorithms when handling large-scale databases. In this study, we introduce QClusformer, a pioneering Transformer-based framework leveraging quantum machines to tackle unsupervised vision clustering challenges. Specifically, we design the Transformer architecture, including the self-attention module and transformer blocks, from a quantum perspective to enable execution on quantum hardware. In addition, we present QClusformer, a variant based on the Transformer architecture, tailored for unsupervised vision clustering tasks. By integrating these elements into an end-to-end framework, QClusformer consistently outperforms previous methods running on classical computers. Empirical evaluations across diverse benchmarks, including MS-Celeb-1M and DeepFashion, underscore the superior performance of QClusformer compared to state-of-the-art methods.

The human brain is a highly efficient processing unit, and understanding how it works can inspire new algorithms and architectures in machine learning. In this work, we introduce a novel framework named Brain Activation Network (BRACTIVE), a transformer-based approach to studying the human visual brain. The main objective of BRACTIVE is to align the visual features of subjects with corresponding brain representations via fMRI signals. It allows us to identify the brain's Regions of Interest (ROI) of the subjects. Unlike previous brain research methods, which can only identify ROIs for one subject at a time and are limited by the number of subjects, BRACTIVE automatically extends this identification to multiple subjects and ROIs. Our experiments demonstrate that BRACTIVE effectively identifies person-specific regions of interest, such as face and body-selective areas, aligning with neuroscience findings and indicating potential applicability to various object categories. More importantly, we found that leveraging human visual brain activity to guide deep neural networks enhances performance across various benchmarks. It encourages the potential of BRACTIVE in both neuroscience and machine intelligence studies.

Topic models have been evolving rapidly over the years, from conventional to recent neural models. However, existing topic models generally struggle with either effectiveness, efficiency, or stability, highly impeding their practical applications. In this paper, we propose FASTopic, a fast, adaptive, stable, and transferable topic model. FASTopic follows a new paradigm: Dual Semantic-relation Reconstruction (DSR). Instead of previous conventional, neural VAE-based or clustering-based methods, DSR discovers latent topics by reconstruction through modeling the semantic relations among document, topic, and word embeddings. This brings about a neat and efficient topic modeling framework. We further propose a novel Embedding Transport Plan (ETP) method. Rather than early straightforward approaches, ETP explicitly regularizes the semantic relations as optimal transport plans. This addresses the relation bias issue and thus leads to effective topic modeling. Extensive experiments on benchmark datasets demonstrate that our FASTopic shows superior effectiveness, efficiency, adaptivity, stability, and transferability, compared to state-of-the-art baselines across various scenarios. Our code is available at https://github.com/bobxwu/FASTopic .

Dynamic topic models track the evolution of topics in sequential documents, which have derived various applications like trend analysis and opinion mining. However, existing models suffer from repetitive topic and unassociated topic issues, failing to reveal the evolution and hindering further applications. To address these issues, we break the tradition of simply chaining topics in existing work and propose a novel neural \modelfullname. We introduce a new evolution-tracking contrastive learning method that builds the similarity relations among dynamic topics. This not only tracks topic evolution but also maintains topic diversity, mitigating the repetitive topic issue. To avoid unassociated topics, we further present an unassociated word exclusion method that consistently excludes unassociated words from discovered topics. Extensive experiments demonstrate our model significantly outperforms state-of-the-art baselines, tracking topic evolution with high-quality topics, showing better performance on downstream tasks, and remaining robust to the hyperparameter for evolution intensities. Our code is available at https://github.com/bobxwu/CFDTM ., Comment: Accepted to ACL 2024 Findings

The problem of detecting spam reviews (opinions) has received significant attention in recent years, especially with the rapid development of e-commerce. Spam reviews are often classified based on comment content, but in some cases, it is insufficient for models to accurately determine the review label. In this work, we introduce the ViSpamReviews v2 dataset, which includes metadata of reviews with the objective of integrating supplementary attributes for spam review classification. We propose a novel approach to simultaneously integrate both textual and categorical attributes into the classification model. In our experiments, the product category proved effective when combined with deep neural network (DNN) models, while text features performed well on both DNN models and the model achieved state-of-the-art performance in the problem of detecting spam reviews on Vietnamese e-commerce websites, namely PhoBERT. Specifically, the PhoBERT model achieves the highest accuracy when combined with product description features generated from the SPhoBert model, which is the combination of PhoBERT and SentenceBERT. Using the macro-averaged F1 score, the task of classifying spam reviews achieved 87.22% (an increase of 1.64% compared to the baseline), while the task of identifying the type of spam reviews achieved an accuracy of 73.49% (an increase of 1.93% compared to the baseline)., Comment: Published in the International Journal of Asian Language Processing (IJALP)

Offline reinforcement learning (RL) addresses the challenge of expensive and high-risk data exploration inherent in RL by pre-training policies on vast amounts of offline data, enabling direct deployment or fine-tuning in real-world environments. However, this training paradigm can compromise policy robustness, leading to degraded performance in practical conditions due to observation perturbations or intentional attacks. While adversarial attacks and defenses have been extensively studied in deep learning, their application in offline RL is limited. This paper proposes a framework to enhance the robustness of offline RL models by leveraging advanced adversarial attacks and defenses. The framework attacks the actor and critic components by perturbing observations during training and using adversarial defenses as regularization to enhance the learned policy. Four attacks and two defenses are introduced and evaluated on the D4RL benchmark. The results show the vulnerability of both the actor and critic to attacks and the effectiveness of the defenses in improving policy robustness. This framework holds promise for enhancing the reliability of offline RL models in practical scenarios.

A fundamental problem in supervised learning is to find a good set of features or distance measures. If the new set of features is of lower dimensionality and can be obtained by a simple transformation of the original data, they can make the model understandable, reduce overfitting, and even help to detect distribution drift. We propose a supervised dimensionality reduction method Gradient Boosting Mapping (GBMAP), where the outputs of weak learners -- defined as one-layer perceptrons -- define the embedding. We show that the embedding coordinates provide better features for the supervised learning task, making simple linear models competitive with the state-of-the-art regressors and classifiers. We also use the embedding to find a principled distance measure between points. The features and distance measures automatically ignore directions irrelevant to the supervised learning task. We also show that we can reliably detect out-of-distribution data points with potentially large regression or classification errors. GBMAP is fast and works in seconds for dataset of million data points or hundreds of features. As a bonus, GBMAP provides a regression and classification performance comparable to the state-of-the-art supervised learning methods., Comment: 32 pages, 8 figures, 5 tables