Search

We introduce the concept of ergodicity and explore its deviation caused by quantum scars in an isolated quantum system, employing a pedagogical approach based on a toy model. Quantum scars, originally identified as traces of classically unstable orbits in certain wavefunctions of chaotic systems, have recently regained interest for their role in non-ergodic dynamics, as they retain memory of their initial states. We elucidate these features of quantum scars within the same framework of this toy model. The integrable part of the model consists of two large spins, with a classical counterpart, which we combine with a random matrix to induce ergodic behavior. Scarred states can be selectively generated from the integrable spin Hamiltonian by protecting them from the ergodic states using a projector method. Deformed projectors mimic the 'quantum leakage' of scarred states, enabling tunable mixing with ergodic states and thereby controlling the degree of scarring. In this simple model, we investigate various properties of quantum scarring and shed light on different aspects of many-body quantum scars observed in more complex quantum systems. Notably, the underlying classicality can be revealed through the entanglement spectrum and the dynamics of 'out-of-time-ordered correlators'.

Course Outcome (CO) and Program Outcome (PO)/Program-Specific Outcome (PSO) alignment is a crucial task for ensuring curriculum coherence and assessing educational effectiveness. The construction of a Course Articulation Matrix (CAM), which quantifies the relationship between COs and POs/PSOs, typically involves assigning numerical values (0, 1, 2, 3) to represent the degree of alignment. In this study, We experiment with four models from the BERT family: BERT Base, DistilBERT, ALBERT, and RoBERTa, and use multiclass classification to assess the alignment between CO and PO/PSO pairs. We first evaluate traditional machine learning classifiers, such as Decision Tree, Random Forest, and XGBoost, and then apply transfer learning to evaluate the performance of the pretrained BERT models. To enhance model interpretability, we apply Explainable AI technique, specifically Local Interpretable Model-agnostic Explanations (LIME), to provide transparency into the decision-making process. Our system achieves accuracy, precision, recall, and F1-score values of 98.66%, 98.67%, 98.66%, and 98.66%, respectively. This work demonstrates the potential of utilizing transfer learning with BERT-based models for the automated generation of CAMs, offering high performance and interpretability in educational outcome assessment., Comment: 26 pages, 9 figures

Vision-brain understanding aims to extract semantic information about brain signals from human perceptions. Existing deep learning methods for vision-brain understanding are usually introduced in a traditional learning paradigm missing the ability to learn the connectivities between brain regions. Meanwhile, the quantum computing theory offers a new paradigm for designing deep learning models. Motivated by the connectivities in the brain signals and the entanglement properties in quantum computing, we propose a novel Quantum-Brain approach, a quantum-inspired neural network, to tackle the vision-brain understanding problem. To compute the connectivity between areas in brain signals, we introduce a new Quantum-Inspired Voxel-Controlling module to learn the impact of a brain voxel on others represented in the Hilbert space. To effectively learn connectivity, a novel Phase-Shifting module is presented to calibrate the value of the brain signals. Finally, we introduce a new Measurement-like Projection module to present the connectivity information from the Hilbert space into the feature space. The proposed approach can learn to find the connectivities between fMRI voxels and enhance the semantic information obtained from human perceptions. Our experimental results on the Natural Scene Dataset benchmarks illustrate the effectiveness of the proposed method with Top-1 accuracies of 95.1% and 95.6% on image and brain retrieval tasks and an Inception score of 95.3% on fMRI-to-image reconstruction task. Our proposed quantum-inspired network brings a potential paradigm to solving the vision-brain problems via the quantum computing theory.

Ensuring a smooth rate of efflux of particles from an outlet without unpredictable clogging events is crucial in processing powders and grains. We show by experiments and simulations that an obstacle placed near the outlet can greatly suppress clog formation in a 3-dimensional granular flow; this counterintuitive phenomenon had previously been demonstrated in 2-dimensional systems. Remarkably, clog suppression is very effective even when the obstacle cross-section is comparable to that of a single grain, and is only a small fraction of the outlet area. Data from two dissimilar obstacle shapes indicate that the underlying mechanism for clog suppression is geometric: while the magnitude of the effect can be affected by dynamical factors, the optimal location of the obstacle is determined entirely by a simple geometric rule. An optimally-placed obstacle arranges for the most-probable clogging arch to be at a perilous and unstable spot, where the local flow geometry is at a point of expansion and a slight perturbation leaves the arch without downstream support. This geometric principle does not rely on previously conjectured dynamical mechanisms and admits generalization to other agent-based and particulate systems.

Latest developments in Large Multimodal Models (LMMs) have broadened their capabilities to include video understanding. Specifically, Text-to-video (T2V) models have made significant progress in quality, comprehension, and duration, excelling at creating videos from simple textual prompts. Yet, they still frequently produce hallucinated content that clearly signals the video is AI-generated. We introduce ViBe: a large-scale Text-to-Video Benchmark of hallucinated videos from T2V models. We identify five major types of hallucination: Vanishing Subject, Numeric Variability, Temporal Dysmorphia, Omission Error, and Physical Incongruity. Using 10 open-source T2V models, we developed the first large-scale dataset of hallucinated videos, comprising 3,782 videos annotated by humans into these five categories. ViBe offers a unique resource for evaluating the reliability of T2V models and provides a foundation for improving hallucination detection and mitigation in video generation. We establish classification as a baseline and present various ensemble classifier configurations, with the TimeSFormer + CNN combination yielding the best performance, achieving 0.345 accuracy and 0.342 F1 score. This benchmark aims to drive the development of robust T2V models that produce videos more accurately aligned with input prompts.

The effectiveness of denoising-driven regularization for image reconstruction has been widely recognized. Two prominent algorithms in this area are Plug-and-Play ($\texttt{PnP}$) and Regularization-by-Denoising ($\texttt{RED}$). We consider two specific algorithms $\texttt{PnP-FISTA}$ and $\texttt{RED-APG}$, where regularization is performed by replacing the proximal operator in the $\texttt{FISTA}$ algorithm with a powerful denoiser. The iterate convergence of $\texttt{FISTA}$ is known to be challenging with no universal guarantees. Yet, we show that for linear inverse problems and a class of linear denoisers, global linear convergence of the iterates of $\texttt{PnP-FISTA}$ and $\texttt{RED-APG}$ can be established through simple spectral analysis.

Mixture-of-Experts (MoE) architectures have recently gained popularity in enabling efficient scaling of large language models. However, we uncover a fundamental tension: while MoEs are designed for selective expert activation, production serving requires request batching, which forces the activation of all experts and negates MoE's efficiency benefits during the decode phase. We present Lynx, a system that enables efficient MoE inference through dynamic, batch-aware expert selection. Our key insight is that expert importance varies significantly across tokens and inference phases, creating opportunities for runtime optimization. Lynx leverages this insight through a lightweight framework that dynamically reduces active experts while preserving model accuracy. Our evaluations show that Lynx achieves up to 1.55x reduction in inference latency while maintaining negligible accuracy loss from baseline model across complex code generation and mathematical reasoning tasks.

As modern web services increasingly rely on REST APIs, their thorough testing has become crucial. Furthermore, the advent of REST API specifications such as the OpenAPI Specification has led to the emergence of many black-box REST API testing tools. However, these tools often focus on individual test elements in isolation (e.g., APIs, parameters, values), resulting in lower coverage and less effectiveness in detecting faults (i.e., 500 response codes). To address these limitations, we present AutoRestTest, the first black-box framework to adopt a dependency-embedded multi-agent approach for REST API testing, integrating Multi-Agent Reinforcement Learning (MARL) with a Semantic Property Dependency Graph (SPDG) and Large Language Models (LLMs). Our approach treats REST API testing as a separable problem, where four agents -- API, dependency, parameter, and value -- collaborate to optimize API exploration. LLMs handle domain-specific value restrictions, the SPDG model simplifies the search space for dependencies using a similarity score between API operations, and MARL dynamically optimizes the agents' behavior. Evaluated on 12 real-world REST services, AutoRestTest outperforms the four leading black-box REST API testing tools, including those assisted by RESTGPT (which augments realistic test inputs using LLMs), in terms of code coverage, operation coverage, and fault detection. Notably, AutoRestTest is the only tool able to identify an internal server error in Spotify. Our ablation study underscores the significant contributions of the agent learning, SPDG, and LLM components., Comment: To be published in the 47th IEEE/ACM International Conference on Software Engineering (ICSE 2025)

Usual parametric and semi-parametric regression methods are inappropriate and inadequate for large clustered survival studies when the appropriate functional forms of the covariates and their interactions in hazard functions are unknown, and random cluster effects and cluster-level covariates are spatially correlated. We present a general nonparametric method for such studies under the Bayesian ensemble learning paradigm called Soft Bayesian Additive Regression Trees. Our methodological and computational challenges include large number of clusters, variable cluster sizes, and proper statistical augmentation of the unobservable cluster-level covariate using a data registry different from the main survival study. We use an innovative 3-step approach based on latent variables to address our computational challenges. We illustrate our method and its advantages over existing methods by assessing the impacts of intervention in some county-level and patient-level covariates to mitigate existing racial disparity in breast cancer survival in 67 Florida counties (clusters) using two different data resources. Florida Cancer Registry (FCR) is used to obtain clustered survival data with patient-level covariates, and the Behavioral Risk Factor Surveillance Survey (BRFSS) is used to obtain further data information on an unobservable county-level covariate of Screening Mammography Utilization (SMU).

This document summarizes the discussions which took place during the PITT-PACC Workshop entitled "Non-Standard Cosmological Epochs and Expansion Histories," held in Pittsburgh, Pennsylvania, Sept. 5-7, 2024. Much like the non-standard cosmological epochs that were the subject of these discussions, the format of this workshop was also non-standard. Rather than consisting of a series of talks from participants, with each person presenting their own work, this workshop was instead organized around free-form discussion blocks, with each centered on a different overall theme and guided by a different set of Discussion Leaders. This document is not intended to serve as a comprehensive review of these topics, but rather as an informal record of the discussions that took place during the workshop, in the hope that the content and free-flowing spirit of these discussions may inspire new ideas and research directions., Comment: 33 pages, LaTeX, 2 figures

Predicting how genetic variation affects phenotypic outcomes at the organismal, cellular, and molecular levels requires deciphering the cis-regulatory code, the sequence rules by which non-coding regions regulate genes. In this perspective, we discuss recent computational progress and challenges towards solving this fundamental problem. We describe how cis-regulatory elements are mapped and how their sequence rules can be learned and interpreted with sequence-to-function neural networks, with the goal of identifying genetic variants in human disease. We also discuss how studies of the 3D chromatin organization could help identifying long-range regulatory effects and how current methods for mapping gene regulatory networks could better describe biological processes. We point out current gaps in knowledge along with technical limitations and benchmarking challenges of computational methods. Finally, we discuss newly emerging technologies, such as spatial transcriptomics, and outline strategies for creating a more general model of the cis-regulatory code that is more broadly applicable across cell types and individuals.

We provide new lower bounds on the privacy guarantee of the multi-epoch Adaptive Batch Linear Queries (ABLQ) mechanism with shuffled batch sampling, demonstrating substantial gaps when compared to Poisson subsampling; prior analysis was limited to a single epoch. Since the privacy analysis of Differentially Private Stochastic Gradient Descent (DP-SGD) is obtained by analyzing the ABLQ mechanism, this brings into serious question the common practice of implementing shuffling-based DP-SGD, but reporting privacy parameters as if Poisson subsampling was used. To understand the impact of this gap on the utility of trained machine learning models, we introduce a practical approach to implement Poisson subsampling at scale using massively parallel computation, and efficiently train models with the same. We compare the utility of models trained with Poisson-subsampling-based DP-SGD, and the optimistic estimates of utility when using shuffling, via our new lower bounds on the privacy guarantee of ABLQ with shuffling., Comment: To appear at NeurIPS 2024

In industry, online randomized controlled experiment (a.k.a A/B experiment) is a standard approach to measure the impact of a causal change. These experiments have small treatment effect to reduce the potential blast radius. As a result, these experiments often lack statistical significance due to low signal-to-noise ratio. To improve the precision (or reduce standard error), we introduce the idea of trigger observations where the output of the treatment and the control model are different. We show that the evaluation with full information about trigger observations (full knowledge) improves the precision in comparison to a baseline method. However, detecting all such trigger observations is a costly affair, hence we propose a sampling based evaluation method (partial knowledge) to reduce the cost. The randomness of sampling introduces bias in the estimated outcome. We theoretically analyze this bias and show that the bias is inversely proportional to the number of observations used for sampling. We also compare the proposed evaluation methods using simulation and empirical data. In simulation, evaluation with full knowledge reduces the standard error as much as 85%. In empirical setup, evaluation with partial knowledge reduces the standard error by 36.48%., Comment: 11 pages, 3 page appendix, 6 figures

The atomic and molecular compounds of cometary ices serve as valuable knowledge into the chemical and physical properties of the outer solar nebula, where comets are formed. From the cometary atmospheres, the atoms and gas-phase molecules arise mainly in three ways: (i) the outgassing from the nucleus, (ii) the photochemical process, and (iii) the sublimation of icy grains from the nucleus. In this paper, we present the radio and millimeter wavelength observation results of Oort cloud non-periodic comet C/2022 E3 (ZTF) using the Giant Metrewave Radio Telescope (GMRT) band L and the Atacama Large Millimeter/Submillimeter Array (ALMA) band 6. We do not detect continuum emissions and an emission line of atomic hydrogen (HI) at rest frequency 1420 MHz from this comet using the GMRT. Based on ALMA observations, we detect the dust continuum emission and rotational emission lines of methanol (CH$_{3}$OH) from comet C/2022 E3 (ZTF). From the dust continuum emission, the activity of dust production (Af$\rho$) of comet ZTF is 2280$\pm$50 cm. Based on LTE spectral modelling, the column density and excitation temperature of CH$_{3}$OH towards C/2022 E3 (ZTF) are (4.50$\pm$0.25)$\times$10$^{14}$ cm$^{-2}$ and 70$\pm$3 K. The integrated emission maps show that CH$_{3}$OH was emitted from the coma region of the comet. The production rate of CH$_{3}$OH towards C/2022 E3 (ZTF) is (7.32$\pm$0.64)$\times$10$^{26}$ molecules s$^{-1}$. The fractional abundance of CH$_{3}$OH with respect to H$_{2}$O in the coma of the comet is 1.52%. We also compare our derived abundance of CH$_{3}$OH with the existence modelled value, and we see the observed and modelled values are nearly similar. We claim that CH$_{3}$OH is formed via the subsequential hydrogenation of formaldehyde (H$_{2}$CO) on the grain surface of comet C/2022 E3 (ZTF)., Comment: Accepted for publication in Research in Astronomy and Astrophysics (RAA)

The Rural Postman Problem with Temporal Unavailability (RPP-TU) is a variant of the Rural Postman Problem (RPP) specified for multi-agent planning over directed graphs with temporal constraints. These temporal constraints represent the unavailable time intervals for each arc during which agents cannot traverse the arc. Such arc unavailability scenarios occur in routing and scheduling of the instrumented wagons for inspection of railway tracks without disturbing the train schedules, i.e. the scheduled trains prohibit access to the signal blocks (sections of railway track separated by signals) for some finite interval of time. A three-index formulation for the RPP-TU is adopted from the literature. The three-index formulation has binary variables for describing the route information of the agents, and continuous non-negative variables to describe the schedules at pre-defined locations. A relaxation of the three-index formulation for RPP-TRU, referred to as Cascaded Graph Formulation (CGF), is investigated in this work. The CGF has attributes that simplify the polyhedral study of time-dependent arc routing problems like RPP-TRU. A novel branch-and-cut algorithm is proposed to solve the RPP-TU, where branching is performed over the service arcs. A family of facet-defining inequalities, derived from the polyhedral study, is used as cutting planes in the proposed branch-and-cut algorithm to reduce the computation time by up to $48\%$. Finally, an application of this work is showcased using a simulation case study of a railway inspection scheduling problem based on Kurla-Vashi-Thane suburban network in Mumbai, India. An improvement of $93\%$ is observed when compared to a Benders' decomposition based MILP solver from the literature., Comment: Visual proof in appendix

We describe a new shadow tomography algorithm that uses $n=\Theta(\sqrt{m}\log m/\epsilon^2)$ samples, for $m$ measurements and additive error $\epsilon$, which is independent of the dimension of the quantum state being learned. This stands in contrast to all previously known algorithms that improve upon the naive approach. The sample complexity also has optimal dependence on $\epsilon$. Additionally, this algorithm is efficient in various aspects, including quantum memory usage (possibly even $O(1)$), gate complexity, classical computation, and robustness to qubit measurement noise. It can also be implemented as a read-once quantum circuit with low quantum memory usage, i.e., it will hold only one copy of $\rho$ in memory, and discard it before asking for a new one, with the additional memory needed being $O(m\log n)$. Our approach builds on the idea of using noisy measurements, but instead of focusing on gentleness in trace distance, we focus on the \textit{gentleness in shadows}, i.e., we show that the noisy measurements do not significantly perturb the expected values., Comment: 29 Pages

In the Plug-and-Play (PnP) method, a denoiser is used as a regularizer within classical proximal algorithms for image reconstruction. It is known that a broad class of linear denoisers can be expressed as the proximal operator of a convex regularizer. Consequently, the associated PnP algorithm can be linked to a convex optimization problem $\mathcal{P}$. For such a linear denoiser, we prove that $\mathcal{P}$ exhibits strong convexity for linear inverse problems. Specifically, we show that the strong convexity of $\mathcal{P}$ can be used to certify objective and iterative convergence of any PnP algorithm derived from classical proximal methods.

Code translation transforms programs from one programming language (PL) to another. Several rule-based transpilers have been designed to automate code translation between different pairs of PLs. However, the rules can become obsolete as the PLs evolve and cannot generalize to other PLs. Recent studies have explored the automation of code translation using Large Language Models (LLMs). One key observation is that such techniques may work well for crafted benchmarks but fail to generalize to the scale and complexity of real-world projects with dependencies, custom types, PL-specific features, etc. We propose AlphaTrans, a neuro-symbolic approach to automate repository-level code translation. AlphaTrans translates both source and test code, and employs multiple levels of validation to ensure the translation preserves the functionality of the source program. To break down the problem for LLMs, AlphaTrans leverages program analysis to decompose the program into fragments and translates them in the reverse call order. We leveraged AlphaTrans to translate ten real-world open-source projects consisting of <836, 8575, 2719> classes, methods, and tests. AlphaTrans translated the entire repository of these projects consisting of 6899 source code fragments. 99.1% of the translated code fragments are syntactically correct, and AlphaTrans validates the translations' runtime behavior and functional correctness for 25.8%. On average, the integrated translation and validation take 36 hours to translate a project, showing its scalability in practice. For the syntactically or semantically incorrect translations, AlphaTrans generates a report including existing translation, stack trace, test errors, or assertion failures. We provided these artifacts to two developers to fix the translation bugs in four projects. They were able to fix the issues in 20.1 hours on average and achieve all passing tests.

We derive a $K$-theoretic analogue of the Vafa--Intriligator formula, computing the (virtual) Euler characteristics of vector bundles over the Quot scheme that compactifies the space of degree $d$ morphisms from a fixed projective curve to the Grassmannian $\mathrm{Gr}(r,N)$. As an application, we deduce interesting vanishing results, used in Part I (arXiv:2406.12191) to study the quantum $K$-ring of $\mathrm{Gr}(r,N)$. In the genus-zero case, we prove a simplified formula involving Schur functions, consistent with the Borel-Weil-Bott theorem in the degree-zero setting. These new formulas offer a novel approach for computing the structure constants of quantum $K$-products., Comment: 45 pages

Mild cognitive impairment (MCI), often linked to early neurodegeneration, is characterized by subtle cognitive declines and disruptions in brain connectivity. The present study offers a detailed analysis of topological changes associated with MCI, focusing on two subtypes: Early MCI and Late MCI. This analysis utilizes fMRI time series data from two distinct populations: the publicly available ADNI dataset (Western cohort) and the in-house TLSA dataset (Indian Urban cohort). Persistent Homology, a topological data analysis method, is employed with two distinct filtration techniques - Vietoris-Rips and graph filtration-for classifying MCI subtypes. For Vietoris-Rips filtration, inter-ROI Wasserstein distance matrices between persistent diagrams are used for classification, while graph filtration relies on the top ten most persistent homology features. Comparative analysis shows that the Vietoris-Rips filtration significantly outperforms graph filtration, capturing subtle variations in brain connectivity with greater accuracy. The Vietoris-Rips filtration method achieved the highest classification accuracy of 85.7\% for distinguishing between age and gender matched healthy controls and MCI, whereas graph filtration reached a maximum accuracy of 71.4\% for the same task. This superior performance highlights the sensitivity of Vietoris-Rips filtration in detecting intricate topological features associated with neurodegeneration. The findings underscore the potential of persistent homology, particularly when combined with the Wasserstein distance, as a powerful tool for early diagnosis and precise classification of cognitive impairments, offering valuable insights into brain connectivity changes in MCI., Comment: 17 pages, 5 figures, 4 tables

Recently, research based on pre-trained models has demonstrated outstanding performance in violence surveillance tasks. However, most of them were black-box systems which faced challenges regarding explainability during training and inference processes. An important question is how to incorporate explicit knowledge into these implicit models, thereby designing expertdriven and interpretable violence surveillance systems. This paper proposes a new paradigm for weakly supervised violence monitoring (WSVM) called Rule base Violence Monitoring (RuleVM). The proposed RuleVM uses a dual-branch structure with different designs for images and text. One of the branches is called the implicit branch, which uses only visual features for coarse-grained binary classification. In this branch, image feature extraction is divided into two channels: one responsible for extracting scene frames and the other focusing on extracting actions. The other branch is called the explicit branch, which utilizes language-image alignment to perform fine-grained classification. For the language channel design in the explicit branch, the proposed RuleVM uses the state-of-the-art YOLOWorld model to detect objects in video frames, and association rules are identified through data mining methods as descriptions of the video. Leveraging the dual-branch architecture, RuleVM achieves interpretable coarse-grained and fine-grained violence surveillance. Extensive experiments were conducted on two commonly used benchmarks, and the results show that RuleVM achieved the best performance in both coarse-grained and finegrained monitoring, significantly outperforming existing state-ofthe-art methods. Moreover, interpretability experiments uncovered some interesting rules, such as the observation that as the number of people increases, the risk level of violent behavior also rises., Comment: 12 pages,7 figures IEEE TSMCA (Under review)

Contextual advertising serves ads that are aligned to the content that the user is viewing. The rapid growth of video content on social platforms and streaming services, along with privacy concerns, has increased the need for contextual advertising. Placing the right ad in the right context creates a seamless and pleasant ad viewing experience, resulting in higher audience engagement and, ultimately, better ad monetization. From a technology standpoint, effective contextual advertising requires a video retrieval system capable of understanding complex video content at a very granular level. Current text-to-video retrieval models based on joint multimodal training demand large datasets and computational resources, limiting their practicality and lacking the key functionalities required for ad ecosystem integration. We introduce ContextIQ, a multimodal expert-based video retrieval system designed specifically for contextual advertising. ContextIQ utilizes modality-specific experts-video, audio, transcript (captions), and metadata such as objects, actions, emotion, etc.-to create semantically rich video representations. We show that our system, without joint training, achieves better or comparable results to state-of-the-art models and commercial solutions on multiple text-to-video retrieval benchmarks. Our ablation studies highlight the benefits of leveraging multiple modalities for enhanced video retrieval accuracy instead of using a vision-language model alone. Furthermore, we show how video retrieval systems such as ContextIQ can be used for contextual advertising in an ad ecosystem while also addressing concerns related to brand safety and filtering inappropriate content., Comment: Accepted at WACV 2025

This paper is about generating motion plans for high degree-of-freedom systems that account for collisions along the entire body. A particular class of mathematical programs with complementarity constraints become useful in this regard. Optimization-based planners can tackle confined-space trajectory planning while being cognizant of robot constraints. However, introducing obstacles in this setting transforms the formulation into a non-convex problem (oftentimes with ill-posed bilinear constraints), which is non-trivial in a real-time setting. To this end, we present the FLIQC (Fast LInear Quadratic Complementarity based) motion planner. Our planner employs a novel motion model that captures the entire rigid robot as well as the obstacle geometry and ensures non-penetration between the surfaces due to the imposed constraint. We perform thorough comparative studies with the state-of-the-art, which demonstrate improved performance. Extensive simulation and hardware experiments validate our claim of generating continuous and reactive motion plans at 1 kHz for modern collaborative robots with constant minimal parameters.

We introduce Planning-guided Retrieval Augmented Generation (Plan$\times$RAG), a novel framework that augments the \emph{retrieve-then-reason} paradigm of existing RAG frameworks to \emph{plan-then-retrieve}. Plan$\times$RAG formulates a reasoning plan as a directed acyclic graph (DAG), decomposing queries into interrelated atomic sub-queries. Answer generation follows the DAG structure, allowing significant gains in efficiency through parallelized retrieval and generation. While state-of-the-art RAG solutions require extensive data generation and fine-tuning of language models (LMs), Plan$\times$RAG incorporates frozen LMs as plug-and-play experts to generate high-quality answers. Compared to existing RAG solutions, Plan$\times$RAG demonstrates significant improvements in reducing hallucinations and bolstering attribution due to its structured sub-query decomposition. Overall, Plan$\times$RAG offers a new perspective on integrating external knowledge in LMs while ensuring attribution by design, contributing towards more reliable LM-based systems., Comment: 22 pages, preprint

With the emergence of low-inertia microgrids powered by inverter-based generation, there remains a concern about the operational resilience of these systems. Grid-forming inverters (GFMs), enabled by various device-level (primary) and system-level (secondary) control methods, are poised to play a significant role in achieving certain operational objectives, such as the effective utilization of clean energy resources while maintaining stability. However, despite the recent advances in GFMs, there is a lack of suitable controls that can ascertain resilience-constrained operations, like maintaining critical operational safety limits during transients under various cyber-physical disruptions. In this work, we develop decentralized autonomous controllers (DACs) that enforce resilience-constrained operation via local, minimally invasive adjustments (e.g., changes in set-points) while co-existing within the hierarchy of existing (primary and secondary) controls. The DACs work autonomously by sensing only local GFM measurements and act only when operational resilience constraints are violated. The proposed DAC scheme is computationally efficient (only algebraic computations), which enables fast, real-time execution and demonstrates the efficacy of the proposed control framework on GridLAB-D-HELICS-based control-grid co-simulations on the IEEE 123-node networked microgrid. Finally, we show how the developed DACs empower the grid by utilizing the available resources entirely to ensure resilience (maintain frequency safe limits)., Comment: This paper is currently under review in a journal

The key components of machine learning are data samples for training, model for learning patterns, and loss function for optimizing accuracy. Analogously, unlearning can potentially be achieved through anti-data samples (or anti-samples), unlearning method, and reversed loss function. While prior research has explored unlearning methods and reversed loss functions, the potential of anti-samples remains largely untapped. In this paper, we introduce UnSTAR: Unlearning with Self-Taught Anti-Sample Reasoning for large language models (LLMs). Our contributions are threefold; first, we propose a novel concept of anti-sample-induced unlearning; second, we generate anti-samples by leveraging misleading rationales, which help reverse learned associations and accelerate the unlearning process; and third, we enable fine-grained targeted unlearning, allowing for the selective removal of specific associations without impacting related knowledge - something not achievable by previous works. Results demonstrate that anti-samples offer an efficient, targeted unlearning strategy for LLMs, opening new avenues for privacy-preserving machine learning and model modification.

Speech anonymisation aims to protect speaker identity by changing personal identifiers in speech while retaining linguistic content. Current methods fail to retain prosody and unique speech patterns found in elderly and pathological speech domains, which is essential for remote health monitoring. To address this gap, we propose a voice conversion-based method (DDSP-QbE) using differentiable digital signal processing and query-by-example. The proposed method, trained with novel losses, aids in disentangling linguistic, prosodic, and domain representations, enabling the model to adapt to uncommon speech patterns. Objective and subjective evaluations show that DDSP-QbE significantly outperforms the voice conversion state-of-the-art concerning intelligibility, prosody, and domain preservation across diverse datasets, pathologies, and speakers while maintaining quality and speaker anonymity. Experts validate domain preservation by analysing twelve clinically pertinent domain attributes., Comment: Accepted in Interspeech 2024

The rising need for explainable deep neural network architectures has utilized semantic concepts as explainable units. Several approaches utilizing disentangled representation learning estimate the generative factors and utilize them as concepts for explaining DNNs. However, even though the generative factors for a dataset remain fixed, concepts are not fixed entities and vary based on downstream tasks. In this paper, we propose a disentanglement mechanism utilizing a variational autoencoder (VAE) for learning mutually independent generative factors for a given dataset and subsequently learning task-specific concepts using a structural causal model (SCM). Our method assumes generative factors and concepts to form a bipartite graph, with directed causal edges from generative factors to concepts. Experiments are conducted on datasets with known generative factors: D-sprites and Shapes3D. On specific downstream tasks, our proposed method successfully learns task-specific concepts which are explained well by the causal edges from the generative factors. Lastly, separate from current causal concept discovery methods, our methodology is generalizable to an arbitrary number of concepts and flexible to any downstream tasks.

Unlike junctions in solid-state devices, a plasma-metal junction (pm-junction) is a junction of classical and quantum electrons. The plasma electrons are Maxwellain in nature, while metal electrons obey the Fermi-Dirac distribution. In this experiment, the current-voltage characteristics of solid-state devices that form homo or hetero-junction are compared to the pm-junction. Observation shows that the turn-on voltage for pn-junction is 0.5V and decreases to 0.24V for metal-semiconductor junction. However, the pm-junction's turn-on voltage was lowered to a negative value of -7.0V. The devices with negative turn-on voltage are suitable for high-frequency operations. Further, observations show that the current-voltage characteristics of the pm-junction depend on the metal's work function, and the turn-on voltage remains unchanged. This result validates the applicability of the energy-band model for the pm-junction. We present a perspective metal-oxide-plasma (MOP), a gaseous electronic device, as an alternative to metal-oxide-semiconductor (MOS), based on the new understanding developed., Comment: Manuscript is of total 9 pages and 10 figures. This manuscript is a fundamental study of junction between an ionised gas i.e. plasma and metal supported by experimental demonstration

We present Movie Gen, a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio. We also show additional capabilities such as precise instruction-based video editing and generation of personalized videos based on a user's image. Our models set a new state-of-the-art on multiple tasks: text-to-video synthesis, video personalization, video editing, video-to-audio generation, and text-to-audio generation. Our largest video generation model is a 30B parameter transformer trained with a maximum context length of 73K video tokens, corresponding to a generated video of 16 seconds at 16 frames-per-second. We show multiple technical innovations and simplifications on the architecture, latent spaces, training objectives and recipes, data curation, evaluation protocols, parallelization techniques, and inference optimizations that allow us to reap the benefits of scaling pre-training data, model size, and training compute for training large scale media generation models. We hope this paper helps the research community to accelerate progress and innovation in media generation models. All videos from this paper are available at https://go.fb.me/MovieGenResearchVideos.

Assessing the quality of bovine satellite cells (BSCs) is essential for the cultivated meat industry, which aims to address global food sustainability challenges. This study aims to develop a label-free method for predicting fluorescence markers in isolated BSCs using deep learning. We employed a U-Net-based CNN model to predict multiple fluorescence signals from a single bright-field microscopy image of cell culture. Two key biomarkers, DAPI and Pax7, were used to determine the abundance and quality of BSCs. The image pre-processing pipeline included fluorescence denoising to improve prediction performance and consistency. A total of 48 biological replicates were used, with statistical performance metrics such as Pearson correlation coefficient and SSIM employed for model evaluation. The model exhibited better performance with DAPI predictions due to uniform staining. Pax7 predictions were more variable, reflecting biological heterogeneity. Enhanced visualization techniques, including color mapping and image overlay, improved the interpretability of the predictions by providing better contextual and perceptual information. The findings highlight the importance of data pre-processing and demonstrate the potential of deep learning to advance non-invasive, label-free assessment techniques in the cultivated meat industry, paving the way for reliable and actionable AI-driven evaluations., Comment: 11 pages, 4 figures

Celikbas, Liang and Sadeghi established a one-sided inequality for the relative version of Jorgensen's dependency formula and questioned whether it would be an equality. In this paper, we show that the inequality can be indeed strict, and prove a relative dependency formula. Along the way, we obtain some bounds on s(M,N), a notion related to the vanishing of relative homology of finitely generated modules M and N over a local ring R, under specific assumptions., Comment: 8 pages

Chart corpora, which comprise data visualizations and their semantic labels, are crucial for advancing visualization research. However, the labels in most existing chart corpora are high-level (e.g., chart types), hindering their utility for broader interactive applications like chart reuse, animation, and accessibility. In this paper, we contribute VisAnatomy, a chart corpus containing 942 real-world SVG charts produced by over 50 tools, encompassing 40 chart types and featuring structural and stylistic design variations. Each chart is augmented with multilevel fine-grained labels on its semantic components, including each graphical element's type, role, and position, hierarchical groupings of elements, group layouts, and visual encodings. We demonstrate the richness of the semantic labels by comparing VisAnatomy with existing corpora. We illustrate the usefulness of VisAnatomy through four applications: chart type classification, chart decomposition, animation authoring, and content navigation for accessibility. Finally, we discuss our plan to improve VisAnatomy and the research opportunities VisAnatomy presents.

An $s$-sparse polynomial has at most $s$ monomials with nonzero coefficients. The Equivalence Testing problem for sparse polynomials (ETsparse) asks to decide if a given polynomial $f$ is equivalent to (i.e., in the orbit of) some $s$-sparse polynomial. In other words, given $f \in \mathbb{F}[\mathbf{x}]$ and $s \in \mathbb{N}$, ETsparse asks to check if there exist $A \in \mathrm{GL}(|\mathbf{x}|, \mathbb{F})$ and $\mathbf{b} \in \mathbb{F}^{|\mathbf{x}|}$ such that $f(A\mathbf{x} + \mathbf{b})$ is $s$-sparse. We show that ETsparse is NP-hard over any field $\mathbb{F}$, if $f$ is given in the sparse representation, i.e., as a list of nonzero coefficients and exponent vectors. This answers a question posed in [Gupta-Saha-Thankey, SODA'23] and [Baraskar-Dewan-Saha, STACS'24]. The result implies that the Minimum Circuit Size Problem (MCSP) is NP-hard for a dense subclass of depth-$3$ arithmetic circuits if the input is given in sparse representation. We also show that approximating the smallest $s_0$ such that a given $s$-sparse polynomial $f$ is in the orbit of some $s_0$-sparse polynomial to within a factor of $s^{\frac{1}{3} - \epsilon}$ is NP-hard for any $\epsilon > 0$; observe that $s$-factor approximation is trivial as the input is $s$-sparse. Finally, we show that for any constant $\sigma \geq 5$, checking if a polynomial (given in sparse representation) is in the orbit of some support-$\sigma$ polynomial is NP-hard. Support of a polynomial $f$ is the maximum number of variables present in any monomial of $f$. These results are obtained via direct reductions from the $3$-SAT problem., Comment: A preliminary version of the paper appeared in the proceedings of ICALP 2024. This version is slightly stronger than the ECCC one: a) We show it is NP-hard to test equivalence to support-5 polynomials (instead of support-6). b) We adapt the proof technique to show it is NP-hard to test equivalence to sparse polynomials under translations only; A suitable gap version of the result is also shown

Large Language Models for Code (or code LLMs) are increasingly gaining popularity and capabilities, offering a wide array of functionalities such as code completion, code generation, code summarization, test generation, code translation, and more. To leverage code LLMs to their full potential, developers must provide code-specific contextual information to the models. These are typically derived and distilled using program analysis tools. However, there exists a significant gap--these static analysis tools are often language-specific and come with a steep learning curve, making their effective use challenging. These tools are tailored to specific program languages, requiring developers to learn and manage multiple tools to cover various aspects of the their code base. Moreover, the complexity of configuring and integrating these tools into the existing development environments add an additional layer of difficulty. This challenge limits the potential benefits that could be gained from more widespread and effective use of static analysis in conjunction with LLMs. To address this challenge, we present codellm-devkit (hereafter, `CLDK'), an open-source library that significantly simplifies the process of performing program analysis at various levels of granularity for different programming languages to support code LLM use cases. As a Python library, CLDK offers developers an intuitive and user-friendly interface, making it incredibly easy to provide rich program analysis context to code LLMs. With this library, developers can effortlessly integrate detailed, code-specific insights that enhance the operational efficiency and effectiveness of LLMs in coding tasks. CLDK is available as an open-source library at https://github.com/IBM/codellm-devkit.

Patterns arise spontaneously in a range of systems spanning the sciences, and their study typically focuses on mechanisms to understand their evolution in space-time. Increasingly, there has been a transition towards controlling these patterns in various functional settings, with implications for engineering. Here, we combine our knowledge of a general class of dynamical laws for pattern formation in non-equilibrium systems, and the power of stochastic optimal control approaches to present a framework that allows us to control patterns at multiple scales, which we dub the "Hamiltonian bridge". We use a mapping between stochastic many-body Lagrangian physics and deterministic Eulerian pattern forming PDEs to leverage our recent approach utilizing the Feynman-Kac-based adjoint path integral formulation for the control of interacting particles and generalize this to the active control of patterning fields. We demonstrate the applicability of our computational framework via numerical experiments on the control of phase separation with and without a conserved order parameter, self-assembly of fluid droplets, coupled reaction-diffusion equations and finally a phenomenological model for spatio-temporal tissue differentiation. We interpret our numerical experiments in terms of a theoretical understanding of how the underlying physics shapes the geometry of the pattern manifold, altering the transport paths of patterns and the nature of pattern interpolation. We finally conclude by showing how optimal control can be utilized to generate complex patterns via an iterative control protocol over pattern forming pdes which can be casted as gradient flows. All together, our study shows how we can systematically build in physical priors into a generative framework for pattern control in non-equilibrium systems across multiple length and time scales., Comment: 29 pages, 8 figures

The multiplicities of positive and negative pions, kaons and unidentified hadrons produced in deep-inelastic scattering are measured in bins of the Bjorken scaling variable $x$, the relative virtual-photon energy $y$ and the fraction of the virtual-photon energy transferred to the final-state hadron $z$. Data were obtained by the COMPASS Collaboration using a 160 GeV muon beam of both electric charges and a liquid hydrogen target. These measurements cover the kinematic domain with photon virtuality $Q^2 > 1$ (GeV/$c)^2$, $0.004 < x < 0.4$, $0.1 < y < 0.7$ and $0.2 < z < 0.85$, in accordance with the kinematic domain used in earlier published COMPASS multiplicity measurements with an isoscalar target. The calculation of radiative corrections was improved by using the Monte Carlo generator DJANGOH, which results in up to 12\% larger corrections in the low-$x$ region., Comment: 19 pages, 29 figures

Heterogeneous graphs are present in various domains, such as social networks, recommendation systems, and biological networks. Unlike homogeneous graphs, heterogeneous graphs consist of multiple types of nodes and edges, each representing different entities and relationships. Generating realistic heterogeneous graphs that capture the complex interactions among diverse entities is a difficult task due to several reasons. The generator has to model both the node type distribution along with the feature distribution for each node type. In this paper, we look into solving challenges in heterogeneous graph generation, by employing a two phase hierarchical structure, wherein the first phase creates a skeleton graph with node types using a prior diffusion based model and in the second phase, we use an encoder and a sampler structure as generator to assign node type specific features to the nodes. A discriminator is used to guide training of the generator and feature vectors are sampled from a node feature pool. We conduct extensive experiments with subsets of IMDB and DBLP datasets to show the effectiveness of our method and also the need for various architecture components.

We give two characterisations of when a map-germ admits a 1-parameter stable unfolding, one related to the $\mathscr K_e$-codimension and another related to the normal form of a versal unfolding. We then prove that there are infinitely many finitely determined map-germs of multiplicity 4 from $\mathbb K^3$ to $\mathbb K^3$ which do not admit a 1-parameter stable unfolding.

The paper presents REPeat, a Real2Sim2Real framework designed to enhance bite acquisition in robot-assisted feeding for soft foods. It uses `pre-acquisition actions' such as pushing, cutting, and flipping to improve the success rate of bite acquisition actions such as skewering, scooping, and twirling. If the data-driven model predicts low success for direct bite acquisition, the system initiates a Real2Sim phase, reconstructing the food's geometry in a simulation. The robot explores various pre-acquisition actions in the simulation, then a Sim2Real step renders a photorealistic image to reassess success rates. If the success improves, the robot applies the action in reality. We evaluate the system on 15 diverse plates with 10 types of food items for a soft food diet, showing improvement in bite acquisition success rates by 27\% on average across all plates. See our project website at https://emprise.cs.cornell.edu/repeat.

Large language models (LLMs) have demonstrated remarkable capabilities; however, the optimization of their prompts has historically prioritized performance metrics at the expense of crucial safety and security considerations. To overcome this shortcoming, we introduce "Survival of the Safest" (SoS), an innovative multi-objective prompt optimization framework that enhances both performance and security in LLMs simultaneously. SoS utilizes an interleaved multi-objective evolution strategy, integrating semantic, feedback, and crossover mutations to effectively traverse the prompt landscape. Differing from the computationally demanding Pareto front methods, SoS provides a scalable solution that expedites optimization in complex, high-dimensional discrete search spaces while keeping computational demands low. Our approach accommodates flexible weighting of objectives and generates a pool of optimized candidates, empowering users to select prompts that optimally meet their specific performance and security needs. Experimental evaluations across diverse benchmark datasets affirm SoS's efficacy in delivering high performance and notably enhancing safety and security compared to single-objective methods. This advancement marks a significant stride towards the deployment of LLM systems that are both high-performing and secure across varied industrial applications, Comment: EMNLP 2024 Industry Track

Machine unlearning algorithms, designed for selective removal of training data from models, have emerged as a promising approach to growing privacy concerns. In this work, we expose a critical yet underexplored vulnerability in the deployment of unlearning systems: the assumption that the data requested for removal is always part of the original training set. We present a threat model where an attacker can degrade model accuracy by submitting adversarial unlearning requests for data not present in the training set. We propose white-box and black-box attack algorithms and evaluate them through a case study on image classification tasks using the CIFAR-10 and ImageNet datasets, targeting a family of widely used unlearning methods. Our results show extremely poor test accuracy following the attack: 3.6% on CIFAR-10 and 0.4% on ImageNet for white-box attacks, and 8.5% on CIFAR-10 and 1.3% on ImageNet for black-box attacks. Additionally, we evaluate various verification mechanisms to detect the legitimacy of unlearning requests and reveal the challenges in verification, as most of the mechanisms fail to detect stealthy attacks without severely impairing their ability to process valid requests. These findings underscore the urgent need for research on more robust request verification methods and unlearning protocols, should the deployment of machine unlearning systems become more prevalent in the future.

Halo assembly bias is a phenomenon whereby the clustering of dark matter halos is dependent on halo properties, such as age, at fixed mass. Understanding the origin of assembly bias is important for interpreting the clustering of galaxies and constraining cosmological models. One proposed explanation for the origin of assembly bias is the truncation of mass accretion in low-mass halos in the presence of more massive halos, called 'arrested development.' Halos undergoing arrested development would have older measured ages and exhibit stronger clustering than equal mass halos that have not undergone arrested development. We propose a new method to test the validity of this explanation for assembly bias, and correct for it in cosmological N-body simulations. The method is based on the idea that the early mass accretion history of a halo, before arrested development takes effect, can be used to predict the late-time evolution of the halo in the absence of arrested development. We implement this idea by fitting a model to the early portion of halo accretion histories and extrapolating to late times. We then calculate 'corrected' masses and ages for halos based on this extrapolation and investigate how this impacts the assembly bias signal. We find that correcting for arrested development this way leads to a factor of two reduction in the strength of the assembly bias signal across a range of low halo masses. This result provides new evidence that arrested development is a cause of assembly bias and validates our approach to mitigating the effect., Comment: 13 pages, 6 figures, accepted for publication in Monthly Notices of the Royal Astronomical Society

Ground motion prediction equations (GMPEs) play a key role in seismic hazard assessment (SHA). Considering the seismo-tectonic, geophysical and geotectonic characteristics of a target region, all the GMPEs may not be suitable in predicting the observed ground motion effectively. With a fairly large number of published GMPEs, the selection and ranking of suitable GMPEs for the design of logic trees in SHA for a particular target region has become a necessity of late. This paper presents a detailed quantitative evaluation of performance of 16 GMPEs against recorded ground motion data in two target regions, characterized by distinct seismo-tectonic, geophysical and geotectonical nature. The data set comprises of 465 three-component spectral accelerograms corresponding to 122 earthquake events. The suitability of a GMPE is tested by two widely accepted data-driven statistical methods, namely, likelihood (LH) and log-likelihood (LLH) method. Different suites of GMPEs are shown suitable for different periods of interest. The results will be useful to scientists and engineers for microzonation and estimation of seismic design parameters for the design of earthquake-resistant structures in these regions., Comment: 24 pages, 5 figures

The possibility of observing correlation-driven charge migration has been a driving force behind theoretical and experimental developments in the field of attosecond molecular science since its inception. Despite significant accomplishments, the unambiguous experimental observation of this quantum beating remains elusive. In this work, we present a method to selectively trigger such dynamics using molecules predicted to exhibit long-lived electron coherence. We show that these dynamics can be selectively triggered using infrared multi-photon ionization and probed using the spacial resolution of X-ray free-electron laser, proposing a promising experimental scheme to study these pivotal dynamics. Additionally, we demonstrate that real-time time-dependent density-functional theory can describe correlation-driven charge migration resulting from a hole mixing structure involving the HOMO of a molecule.

Generalized additive models (GAMs) have long been a powerful white-box tool for the intelligible analysis of tabular data, revealing the influence of each feature on the model predictions. Despite the success of neural networks (NNs) in various domains, their application as NN-based GAMs in tabular data analysis remains suboptimal compared to tree-based ones, and the opacity of encoders in NN-GAMs also prevents users from understanding how networks learn the functions. In this work, we propose a new deep tabular learning method, termed Prototypical Neural Additive Model (ProtoNAM), which introduces prototypes into neural networks in the framework of GAMs. With the introduced prototype-based feature activation, ProtoNAM can flexibly model the irregular mapping from tabular features to the outputs while maintaining the explainability of the final prediction. We also propose a gradient-boosting inspired hierarchical shape function modeling method, facilitating the discovery of complex feature patterns and bringing transparency into the learning process of each network layer. Our empirical evaluations demonstrate that ProtoNAM outperforms all existing NN-based GAMs, while providing additional insights into the shape function learned for each feature. The source code of ProtoNAM is available at \url{https://github.com/Teddy-XiongGZ/ProtoNAM}.