Search

A de Bruijn sequence of order $k$ over a finite alphabet is a cyclic sequence with the property that it contains every possible $k$-sequence as a substring exactly once. Orthogonal de Bruijn sequences are collections of de Bruijn sequences of the same order, $k$, satisfying the joint constraint that every $(k+1)$-sequence appears as a substring in at most one of the sequences in the collection. Both de Bruijn and orthogonal de Bruijn sequences have found numerous applications in synthetic biology, although the latter topic remains largely unexplored in the coding theory literature. Here we study three relevant practical generalizations of orthogonal de Bruijn sequences where we relax either the constraint that every $(k+1)$-sequence appears exactly once, or that the sequences themselves are de Bruijn rather than balanced de Bruijn sequences. We also provide lower and upper bounds on the number of fixed-weight orthogonal de Bruijn sequences., Comment: 7 pages, 3 figures

We have developed and tested a simplified but versatile numerical model of nanoparticles' aggregation using Langevin dynamics. The model is particularly capable of simulating aggregation in an evaporating (or condensing) microdroplet. It runs on a graphics processing unit (GPU), which makes it sufficiently fast for real-time conceptualization tasks. We have verified the results of modeling against the findings from two types of experiments we conducted in electrodynamic traps. Firstly, our model helped us to elucidate the phenomenon of scattering `revival', often observed during the evaporation of composite microdroplets. Further on, we were able to mimic our experiments, in which the microdroplets were dried up to form nanoparticle (NP) aggregates, and then soft-landed. Thus we could compare model predictions with SEM imaging. The model was tested for up to $2.5\times 10^5$ nanoparticles of several coexisting types. Several types of interactions can be accounted for: inter-particle: Lennard-Jones and Coulomb; external: dispersion medium viscosity, centrifugal force, gravity, surface tension, and interface movement. Brownian motion of nanoparticles can be freely controlled. The core program is accompanied by scripts extracting statistical NP aggregates properties in post-processing -- fractal dimension and radial distribution functions. The codes are made available in public repositories. Several diverse evolution scenarios are presented.

Recent advancements in Vision-Language Models (VLMs) have sparked interest in their use for autonomous driving, particularly in generating interpretable driving decisions through natural language. However, the assumption that VLMs inherently provide visually grounded, reliable, and interpretable explanations for driving remains largely unexamined. To address this gap, we introduce DriveBench, a benchmark dataset designed to evaluate VLM reliability across 17 settings (clean, corrupted, and text-only inputs), encompassing 19,200 frames, 20,498 question-answer pairs, three question types, four mainstream driving tasks, and a total of 12 popular VLMs. Our findings reveal that VLMs often generate plausible responses derived from general knowledge or textual cues rather than true visual grounding, especially under degraded or missing visual inputs. This behavior, concealed by dataset imbalances and insufficient evaluation metrics, poses significant risks in safety-critical scenarios like autonomous driving. We further observe that VLMs struggle with multi-modal reasoning and display heightened sensitivity to input corruptions, leading to inconsistencies in performance. To address these challenges, we propose refined evaluation metrics that prioritize robust visual grounding and multi-modal understanding. Additionally, we highlight the potential of leveraging VLMs' awareness of corruptions to enhance their reliability, offering a roadmap for developing more trustworthy and interpretable decision-making systems in real-world autonomous driving contexts. The benchmark toolkit is publicly accessible., Comment: Preprint; 41 pages, 32 figures, 16 tables; Project Page at https://drive-bench.github.io/

To make efficient use of limited physical resources, the brain must match its coding and computational strategies to the statistical structure of input signals. An attractive testing ground for these principles is the problem of motion estimation in the fly visual system: we understand the optics of the compound eye, have a quantitative description of input signals and noise from the retina, and can record from output neurons that encode estimates of different velocity components. Furthermore, recent work provides a nearly complete wiring diagram of the intervening circuitry. What is missing is a characterization of the visual signals and motions that flies encounter in a natural context. We attack this directly with the development of a specialized camera that matches the high temporal resolution, optical properties, and spectral sensitivity of the fly's eye; inertial motion sensors provide ground truth about rotations and translations through the world. We describe the design, construction, and performance characteristics of this FlEye camera. To illustrate the opportunities created by this instrument we use data on movies and motion to construct optimal local motion estimators that can be compared with the responses of the fly's motion sensitive neurons.

Signed Graph Neural Networks (SGNNs) are powerful tools for signed graph representation learning but struggle with limited generalization and heavy dependence on labeled data. While recent advancements in "graph pre-training and prompt tuning" have reduced label dependence in Graph Neural Networks (GNNs) and improved their generalization abilities by leveraging pre-training knowledge, these efforts have focused exclusively on unsigned graphs. The scarcity of publicly available signed graph datasets makes it essential to transfer knowledge from unsigned graphs to signed graph tasks. However, this transfer introduces significant challenges due to the graph-level and task-level divergences between the pre-training and downstream phases. To address these challenges, we propose Signed Graph Prompt Tuning (SGPT) in this paper. Specifically, SGPT employs a graph template and a semantic prompt to segregate mixed link semantics in the signed graph and then adaptively integrate the distinctive semantic information according to the needs of downstream tasks, thereby unifying the pre-training and downstream graphs. Additionally, SGPT utilizes a task template and a feature prompt to reformulate the downstream signed graph tasks, aligning them with pre-training tasks to ensure a unified optimization objective and consistent feature space across tasks. Finally, extensive experiments are conducted on popular signed graph datasets, demonstrating the superiority of SGPT over state-of-the-art methods.

NGC 5128 (Centaurus A), the closest giant elliptical galaxy outside the Local Group to the Milky Way, is one of the brightest extragalactic radio sources. It is distinguished by a prominent dust lane and powerful jets, driven by a supermassive black hole at its core. Using previously identified long-period variable (LPV) stars from the literature, this study aims to reconstruct the star formation history (SFH) of two distinct regions in the halo of NGC 5128. These regions reveal remarkably similar SFHs, despite being located about 28 kpc apart on opposite sides of the galaxy's center. In Field 1, star formation rates (SFRs) show notable increases at approximately 800 Myr and 3.8 Gyr ago. Field 2 exhibits similar peaks at these times, along with an additional rise around 6.3 Gyr ago. The increase in SFR around 800 Myr ago is consistent with earlier research suggesting a merger event. Since no LPV catalog exists for the central region of NGC 5128, we focused our investigation on its outer regions, which has provided new insights into the complex evolutionary history of this cornerstone galaxy. The SFH traced by LPVs supports a scenario in which multiple events of nuclear activity have triggered episodic, jet-induced star formation., Comment: 6 pages, 4 figures, To be published in Communications of BAO (ComBAO)

Functional magnetic resonance imaging (fMRI) is extensively used in clinical and preclinical settings to study brain function, however, fMRI data is inherently noisy due to physiological processes, hardware, and external noise. Denoising is one of the main preprocessing steps in any fMRI analysis pipeline. This process is challenging in preclinical data in comparison to clinical data due to variations in brain geometry, image resolution, and low signal-to-noise ratios. In this paper, we propose a structure-preserved algorithm based on a 3D Wasserstein generative adversarial network with a 3D dense U-net based discriminator called, 3D U-WGAN. We apply a 4D data configuration to effectively denoise temporal and spatial information in analyzing preclinical fMRI data. GAN-based denoising methods often utilize a discriminator to identify significant differences between denoised and noise-free images, focusing on global or local features. To refine the fMRI denoising model, our method employs a 3D dense U-Net discriminator to learn both global and local distinctions. To tackle potential over-smoothing, we introduce an adversarial loss and enhance perceptual similarity by measuring feature space distances. Experiments illustrate that 3D U-WGAN significantly improves image quality in resting-state and task preclinical fMRI data, enhancing signal-to-noise ratio without introducing excessive structural changes in existing methods. The proposed method outperforms state-of-the-art methods when applied to simulated and real data in a fMRI analysis pipeline.

As an intriguing case is the goodness of the machine and deep learning models generated by these LLMs in conducting automated scientific data analysis, where a data analyst may not have enough expertise in manually coding and optimizing complex deep learning models and codes and thus may opt to leverage LLMs to generate the required models. This paper investigates and compares the performance of the mainstream LLMs, such as ChatGPT, PaLM, LLama, and Falcon, in generating deep learning models for analyzing time series data, an important and popular data type with its prevalent applications in many application domains including financial and stock market. This research conducts a set of controlled experiments where the prompts for generating deep learning-based models are controlled with respect to sensitivity levels of four criteria including 1) Clarify and Specificity, 2) Objective and Intent, 3) Contextual Information, and 4) Format and Style. While the results are relatively mix, we observe some distinct patterns. We notice that using LLMs, we are able to generate deep learning-based models with executable codes for each dataset seperatly whose performance are comparable with the manually crafted and optimized LSTM models for predicting the whole time series dataset. We also noticed that ChatGPT outperforms the other LLMs in generating more accurate models. Furthermore, we observed that the goodness of the generated models vary with respect to the ``temperature'' parameter used in configuring LLMS. The results can be beneficial for data analysts and practitioners who would like to leverage generative AIs to produce good prediction models with acceptable goodness.

Pruning neural networks, which involves removing a fraction of their weights, can often maintain high accuracy while significantly reducing model complexity, at least up to a certain limit. We present a neural network pruning technique that builds upon the Combinatorial Brain Surgeon, but solves an optimization problem over a subset of the network weights in an iterative, block-wise manner using block coordinate descent. The iterative, block-based nature of this pruning technique, which we dub ``iterative Combinatorial Brain Surgeon'' (iCBS) allows for scalability to very large models, including large language models (LLMs), that may not be feasible with a one-shot combinatorial optimization approach. When applied to large models like Mistral and DeiT, iCBS achieves higher performance metrics at the same density levels compared to existing pruning methods such as Wanda. This demonstrates the effectiveness of this iterative, block-wise pruning method in compressing and optimizing the performance of large deep learning models, even while optimizing over only a small fraction of the weights. Moreover, our approach allows for a quality-time (or cost) tradeoff that is not available when using a one-shot pruning technique alone. The block-wise formulation of the optimization problem enables the use of hardware accelerators, potentially offsetting the increased computational costs compared to one-shot pruning methods like Wanda. In particular, the optimization problem solved for each block is quantum-amenable in that it could, in principle, be solved by a quantum computer., Comment: 16 pages, 6 figures, 5 tables

Conformal prediction (CP) is a distribution-free framework for achieving probabilistic guarantees on black-box models. CP is generally applied to a model post-training. Recent research efforts, on the other hand, have focused on optimizing CP efficiency during training. We formalize this concept as the problem of conformal risk minimization (CRM). In this direction, conformal training (ConfTr) by Stutz et al.(2022) is a technique that seeks to minimize the expected prediction set size of a model by simulating CP in-between training updates. Despite its potential, we identify a strong source of sample inefficiency in ConfTr that leads to overly noisy estimated gradients, introducing training instability and limiting practical use. To address this challenge, we propose variance-reduced conformal training (VR-ConfTr), a CRM method that incorporates a variance reduction technique in the gradient estimation of the ConfTr objective function. Through extensive experiments on various benchmark datasets, we demonstrate that VR-ConfTr consistently achieves faster convergence and smaller prediction sets compared to baselines.

Data deduplication, one of the key features of modern Big Data storage devices, is the process of removing replicas of data chunks stored by different users. Despite the importance of deduplication, several drawbacks of the method, such as storage robustness and file fragmentation, have not been previously analyzed from a theoretical point of view. Storage robustness pertains to ensuring that deduplicated data can be used to reconstruct the original files without service disruptions and data loss. Fragmentation pertains to the problems of placing deduplicated data chunks of different user files in a proximity-preserving linear order, since neighboring chunks of the same file may be stored in sectors far apart on the server. This work proposes a new theoretical model for data fragmentation and introduces novel graph- and coding-theoretic approaches for reducing fragmentation via limited duplication (repetition coding) and coded deduplication (e.g., linear coding). In addition to alleviating issues with fragmentation, limited duplication and coded deduplication can also serve the dual purpose of increasing the robusteness of the system design. The contributions of our work are three-fold. First, we describe a new model for file structures in the form of self-avoiding (simple) paths in specialized graphs. Second, we introduce several new metrics for measuring the fragmentation level in deduplication systems on graph-structured files, including the stretch metric that captures the worst-case "spread" of adjacent data chunks within a file when deduplicated and placed on the server; and, the jump metric that captures the worst-case number of times during the reconstruction process of a file that one has to change the readout location on the server. For the stretch metric, we establish a connection between the level of fragmentation and the bandwidth of the file-graph. In particular, ...

We introduce a Retrieval-Augmented Generation (RAG) system for translating user questions into accurate federated SPARQL queries over bioinformatics knowledge graphs (KGs) leveraging Large Language Models (LLMs). To enhance accuracy and reduce hallucinations in query generation, our system utilises metadata from the KGs, including query examples and schema information, and incorporates a validation step to correct generated queries. The system is available online at chat.expasy.org.

Background. In the last decades, several life science resources have structured data using the same framework and made these accessible using the same query language to facilitate interoperability. Knowledge graphs have seen increased adoption in bioinformatics due to their advantages for representing data in a generic graph format. For example, yummydata.org catalogs more than 60 knowledge graphs accessible through SPARQL, a technical query language. Although SPARQL allows powerful, expressive queries, even across physically distributed knowledge graphs, formulating such queries is a challenge for most users. Therefore, to guide users in retrieving the relevant data, many of these resources provide representative examples. These examples can also be an important source of information for machine learning, if a sufficiently large number of examples are provided and published in a common, machine-readable and standardized format across different resources. Findings. We introduce a large collection of human-written natural language questions and their corresponding SPARQL queries over federated bioinformatics knowledge graphs (KGs) collected for several years across different research groups at the SIB Swiss Institute of Bioinformatics. The collection comprises more than 1000 example questions and queries, including 65 federated queries. We propose a methodology to uniformly represent the examples with minimal metadata, based on existing standards. Furthermore, we introduce an extensive set of open-source applications, including query graph visualizations and smart query editors, easily reusable by KG maintainers who adopt the proposed methodology. Conclusions. We encourage the community to adopt and extend the proposed methodology, towards richer KG metadata and improved Semantic Web services.

End-effector based assistive robots face persistent challenges in generating smooth and robust trajectories when controlled by human's noisy and unreliable biosignals such as muscle activities and brainwaves. The produced endpoint trajectories are often jerky and imprecise to perform complex tasks such as stable robotic grasping. We propose STREAMS (Self-Training Robotic End-to-end Adaptive Multimodal Shared autonomy) as a novel framework leveraged deep reinforcement learning to tackle this challenge in biosignal based robotic control systems. STREAMS blends environmental information and synthetic user input into a Deep Q Learning Network (DQN) pipeline for an interactive end-to-end and self-training mechanism to produce smooth trajectories for the control of end-effector based robots. The proposed framework achieved a high-performance record of 98% in simulation with dynamic target estimation and acquisition without any pre-existing datasets. As a zero-shot sim-to-real user study with five participants controlling a physical robotic arm with noisy head movements, STREAMS (as an assistive mode) demonstrated significant improvements in trajectory stabilization, user satisfaction, and task performance reported as a success rate of 83% compared to manual mode which was 44% without any task support. STREAMS seeks to improve biosignal based assistive robotic controls by offering an interactive, end-to-end solution that stabilizes end-effector trajectories, enhancing task performance and accuracy.

Distributed quantum computing (DQC) has emerged as a promising approach to overcome the scalability limitations of monolithic quantum processors in terms of computing capability. However, realising the full potential of DQC requires effective resource allocation. This involves efficiently distributing quantum circuits across the network by assigning each circuit to an optimal subset of quantum processing units (QPUs), based on factors such as their computational power and connectivity. In heterogeneous DQC networks with arbitrary topologies and non-identical QPUs, resource allocation becomes a complex challenge. This paper addresses the problem of resource allocation in such networks, focusing on computing resource management in a quantum farm setting. We propose a multi-objective optimisation algorithm for optimal QPU allocation that aims to minimise the degradation caused by inter-QPU communication latencies due to qubit decoherence, while maximising the number of concurrently assignable quantum circuits. The algorithm takes into account several key factors, including the network topology, QPU characteristics, and quantum circuit structure, to make efficient allocation decisions. We employ mixed integer linear programming to solve this optimisation problem. Simulation results demonstrate the effectiveness of the proposed algorithm in minimising communication costs and improving resource utilisation compared to a benchmark greedy allocation approach. Notably, assuming a single circuit partition per QPU, the success rate of quantum circuit assignments improves by 5.25%-13.75%. To complement our proposed QPU allocation method, we also present a compatible quantum circuit scheduling model. Our work provides valuable insights into resource allocation strategies for DQC systems and contributes to the development of efficient execution management frameworks for quantum computing., Comment: 12 pages, 9 figures

End-to-end architectures in autonomous driving (AD) face a significant challenge in interpretability, impeding human-AI trust. Human-friendly natural language has been explored for tasks such as driving explanation and 3D captioning. However, previous works primarily focused on the paradigm of declarative interpretability, where the natural language interpretations are not grounded in the intermediate outputs of AD systems, making the interpretations only declarative. In contrast, aligned interpretability establishes a connection between language and the intermediate outputs of AD systems. Here we introduce Hint-AD, an integrated AD-language system that generates language aligned with the holistic perception-prediction-planning outputs of the AD model. By incorporating the intermediate outputs and a holistic token mixer sub-network for effective feature adaptation, Hint-AD achieves desirable accuracy, achieving state-of-the-art results in driving language tasks including driving explanation, 3D dense captioning, and command prediction. To facilitate further study on driving explanation task on nuScenes, we also introduce a human-labeled dataset, Nu-X. Codes, dataset, and models will be publicly available., Comment: CoRL 2024, Project Page: https://air-discover.github.io/Hint-AD/

Rank aggregation combines multiple ranked lists into a consensus ranking. In fields like biomedical data sharing, rankings may be distributed and require privacy. This motivates the need for federated rank aggregation protocols, which support distributed, private, and communication-efficient learning across multiple clients with local data. We present the first known federated rank aggregation methods using Borda scoring and Lehmer codes, focusing on the sample complexity for federated algorithms on Mallows distributions with a known scaling factor $\phi$ and an unknown centroid permutation $\sigma_0$. Federated Borda approach involves local client scoring, nontrivial quantization, and privacy-preserving protocols. We show that for $\phi \in [0,1)$, and arbitrary $\sigma_0$ of length $N$, it suffices for each of the $L$ clients to locally aggregate $\max\{C_1(\phi), C_2(\phi)\frac{1}{L}\log \frac{N}{\delta}\}$ rankings, where $C_1(\phi)$ and $C_2(\phi)$ are constants, quantize the result, and send it to the server who can then recover $\sigma_0$ with probability $\geq 1-\delta$. Communication complexity scales as $NL \log N$. Our results represent the first rigorous analysis of Borda's method in centralized and distributed settings under the Mallows model. Federated Lehmer coding approach creates a local Lehmer code for each client, using a coordinate-majority aggregation approach with specialized quantization methods for efficiency and privacy. We show that for $\phi+\phi^2<1+\phi^N$, and arbitrary $\sigma_0$ of length $N$, it suffices for each of the $L$ clients to locally aggregate $\max\{C_3(\phi), C_4(\phi)\frac{1}{L}\log \frac{N}{\delta}\}$ rankings, where $C_3(\phi)$ and $C_4(\phi)$ are constants. Clients send truncated Lehmer coordinate histograms to the server, which can recover $\sigma_0$ with probability $\geq 1-\delta$. Communication complexity is $\sim O(N\log NL\log L)$.

There is evidence that magnetism can potentially increase the thermopower of materials, most likely due to magnon scattering, suggesting the incorporation of intrinsic magnetic semiconductors in non-magnetic thermoelectric materials. Here, samples of $\textit{p}$-type Bi$_{0.5}$Sb$_{1.5}$Te$_{3}$ with 10 at.% excess Te are ball-milled with varying ratio of the antiferromagnetic semiconductor CrSb (0, 0.125, 0.5, and 1 wt.%) to prepare bulk samples by spark plasma sintering technique. The thermopower of samples containing CrSb is increased due to an increase in the effective mass of the charge carriers, indicating that there is a drag effect originating from the magnetic particles. However, this was at the expense of reduced electrical conductivity caused by reduced charge carrier mobility. While overall only marginal improvements in power factors were observed, these samples exhibited significantly lower thermal conductivity compared to the single-phase material. As a result, a peak $\textit{zT}$ value of $\sim$1.4 was achieved at 325 K for the sample with 0.125 wt.% CrSb. These results highlight the potential of incorporating magnetic secondary phases to enhance the thermoelectric performance of materials.

Reward shaping is effective in addressing the sparse-reward challenge in reinforcement learning by providing immediate feedback through auxiliary informative rewards. Based on the reward shaping strategy, we propose a novel multi-task reinforcement learning framework, that integrates a centralized reward agent (CRA) and multiple distributed policy agents. The CRA functions as a knowledge pool, which aims to distill knowledge from various tasks and distribute it to individual policy agents to improve learning efficiency. Specifically, the shaped rewards serve as a straightforward metric to encode knowledge. This framework not only enhances knowledge sharing across established tasks but also adapts to new tasks by transferring valuable reward signals. We validate the proposed method on both discrete and continuous domains, demonstrating its robustness in multi-task sparse-reward settings and its effective transferability to unseen tasks.

Reward shaping is a technique in reinforcement learning that addresses the sparse-reward problem by providing more frequent and informative rewards. We introduce a self-adaptive and highly efficient reward shaping mechanism that incorporates success rates derived from historical experiences as shaped rewards. The success rates are sampled from Beta distributions, which dynamically evolve from uncertain to reliable values as data accumulates. Initially, the shaped rewards exhibit more randomness to encourage exploration, while over time, the increasing certainty enhances exploitation, naturally balancing exploration and exploitation. Our approach employs Kernel Density Estimation (KDE) combined with Random Fourier Features (RFF) to derive the Beta distributions, providing a computationally efficient, non-parametric, and learning-free solution for high-dimensional continuous state spaces. Our method is validated on various tasks with extremely sparse rewards, demonstrating notable improvements in sample efficiency and convergence stability over relevant baselines.

Fully supervised training of semantic segmentation models is costly and challenging because each pixel within an image needs to be labeled. Therefore, the sparse pixel-level annotation methods have been introduced to train models with a subset of pixels within each image. We introduce a Bayesian active learning framework based on sparse pixel-level annotation that utilizes a pixel-level Bayesian uncertainty measure based on Balanced Entropy (BalEnt) [84]. BalEnt captures the information between the models' predicted marginalized probability distribution and the pixel labels. BalEnt has linear scalability with a closed analytical form and can be calculated independently per pixel without relational computations with other pixels. We train our proposed active learning framework for Cityscapes, Camvid, ADE20K and VOC2012 benchmark datasets and show that it reaches supervised levels of mIoU using only a fraction of labeled pixels while outperforming the previous state-of-the-art active learning models with a large margin.

Due to the expected disparity in quantum vs. classical clock speeds, quantum advantage for branch and bound algorithms is more likely achievable in settings involving large search trees and low operator evaluation costs. Therefore, in this paper, we describe and experimentally validate an exact classical branch and bound solver for quadratic unconstrained binary optimization (QUBO) problems that matches these criteria. Our solver leverages cheap-to-implement bounds from the literature previously proposed for Ising models, including that of Hartwig, Daske, and Kobe from 1984. We detail a variety of techniques from high-performance computing and operations research used to boost solver performance, including a global variable reordering heuristic, a primal heuristic based on simulated annealing, and a truncated computation of the recursive bound. We also outline a number of simple and inexpensive bound extrapolation techniques. Finally, we conduct an extensive empirical analysis of our solver, comparing its performance to state-of-the-art QUBO and MaxCut solvers, and discuss the challenges of a speedup via quantum branch and bound beyond those faced by any quadratic quantum speedup.

With recent advancements in AI and computation tools, intelligent paradigms emerged to empower different fields such as healthcare robots with new capabilities. Advanced AI robotic algorithms (e.g., reinforcement learning) can be trained and developed to autonomously make individual decisions to achieve a desired and usually fixed goal. However, such independent decisions and goal achievements might not be ideal for a healthcare robot that usually interacts with a dynamic end-user or a patient. In such a complex human-robot interaction (teaming) framework, the dynamic user continuously wants to be involved in decision-making as well as introducing new goals while interacting with their present environment in real-time. To address this challenge, an adaptive shared autonomy AI paradigm is required to be developed for the two interactive agents (Human & AI agents) with a foundation based on human-centered factors to avoid any possible ethical issues and guarantee no harm to humanity., Comment: 5 pages, 1 figure

A combinatorial trade is a pair of sets of blocks of elements that can be exchanged while preserving relevant subset intersection constraints. The class of balanced and swap-robust minimal trades was proposed in [1] for exchanging blocks of data chunks stored on distributed storage systems in an access- and load-balanced manner. More precisely, data chunks in the trades of interest are labeled by popularity ranks and the blocks are required to have both balanced overall popularity and stability properties with respect to swaps in chunk popularities. The original construction of such trades relied on computer search and paired balanced sets obtained through iterative combining of smaller sets that have provable stability guarantees. To reduce the substantial gap between the results of prior approaches and the known theoretical lower bound, we present new analytical upper and lower bounds on the minimal disbalance of blocks introduced by limited-magnitude popularity ranking swaps. Our constructive and near-optimal approach relies on pairs of graphs whose vertices are two balanced sets with edges/arcs that capture the balance and potential balance changes induced by limited-magnitude popularity swaps. In particular, we show that if we start with carefully selected balanced trades and limit the magnitude of rank swaps to one, the new upper and lower bound on the maximum block disbalance caused by a swap only differ by a factor of $1.07$. We also extend these results for larger popularity swap magnitudes., Comment: arXiv admin note: text overlap with arXiv:2303.12996

Recurrent neural network-based sequence-to-sequence models have been extensively applied for multi-step-ahead time series forecasting. These models typically involve a decoder trained using either its previous forecasts or the actual observed values as the decoder inputs. However, relying on self-generated predictions can lead to the rapid accumulation of errors over multiple steps, while using the actual observations introduces exposure bias as these values are unavailable during the extrapolation stage. In this regard, this study proposes a novel training approach called reinforced decoder, which introduces auxiliary models to generate alternative decoder inputs that remain accessible when extrapolating. Additionally, a reinforcement learning algorithm is utilized to dynamically select the optimal inputs to improve accuracy. Comprehensive experiments demonstrate that our approach outperforms representative training methods over several datasets. Furthermore, the proposed approach also exhibits promising performance when generalized to self-attention-based sequence-to-sequence forecasting models., Comment: 12 pages,8 figures

The open-vocabulary image segmentation task involves partitioning images into semantically meaningful segments and classifying them with flexible text-defined categories. The recent vision-based foundation models such as the Segment Anything Model (SAM) have shown superior performance in generating class-agnostic image segments. The main challenge in open-vocabulary image segmentation now lies in accurately classifying these segments into text-defined categories. In this paper, we introduce the Universal Segment Embedding (USE) framework to address this challenge. This framework is comprised of two key components: 1) a data pipeline designed to efficiently curate a large amount of segment-text pairs at various granularities, and 2) a universal segment embedding model that enables precise segment classification into a vast range of text-defined categories. The USE model can not only help open-vocabulary image segmentation but also facilitate other downstream tasks (e.g., querying and ranking). Through comprehensive experimental studies on semantic segmentation and part segmentation benchmarks, we demonstrate that the USE framework outperforms state-of-the-art open-vocabulary segmentation methods.

NGC 5128 (Cen A) is the nearest giant elliptical galaxy and one of the brightest extragalactic radio sources in the sky, boasting a prominent dust lane and jets emanating from its nuclear supermassive black hole. In this paper, we construct the star formation history (SFH) of two small fields in the halo of NGC 5128: a northeastern field (Field 1) at a projected distance of $\sim 18.8$ kpc from the center, and a southern field (Field 2) $\sim 9.9$ kpc from the center. Our method is based on identifying long period variable (LPV) stars that trace their sibling stellar population and hence historical star formation due to their high luminosity and strong variability; we identified 395 LPVs in Field 1 and 671 LPVs in Field 2. Even though the two fields are $\sim 28$ kpc apart on opposite sides from the center, they show similar SFHs. In Field 1, star formation rates (SFRs) increased significantly around $t\sim 800$ Myr and $t\sim 3.8$ Gyr; and in Field 2, SFRs increased considerably around $t\sim 800$ Myr, $t\sim 3.8$ Gyr, and $t\sim 6.3$ Gyr, where $t$ is look--back time. The increase in SFR $\sim 800$ Myr ago agrees with previous suggestions that the galaxy experienced a merger around that time. The SFH reconstructed from LPVs supports a scenario in which multiple episodes of nuclear activity lead to episodic jet-induced star formation. While there is no catalog of LPVs for the central part of NGC 5128, applying our method to the outer regions (for the first time in a galaxy outside the Local Group) has enabled us to put constraints on the complex evolution of this cornerstone galaxy., Comment: Accepted for publication in ApJ

Large language models (LLMs) have attracted considerable attention as they are capable of showcasing impressive capabilities generating comparable high-quality responses to human inputs. LLMs, can not only compose textual scripts such as emails and essays but also executable programming code. Contrary, the automated reasoning capability of these LLMs in performing statistically-driven descriptive analysis, particularly on user-specific data and as personal assistants to users with limited background knowledge in an application domain who would like to carry out basic, as well as advanced statistical and domain-specific analysis is not yet fully explored. More importantly, the performance of these LLMs has not been compared and discussed in detail when domain-specific data analysis tasks are needed. This study, consequently, explores whether LLMs can be used as generative AI-based personal assistants to users with minimal background knowledge in an application domain infer key data insights. To demonstrate the performance of the LLMs, the study reports a case study through which descriptive statistical analysis, as well as Natural Language Processing (NLP) based investigations, are performed on a number of phishing emails with the objective of comparing the accuracy of the results generated by LLMs to the ones produced by analysts. The experimental results show that LangChain and the Generative Pre-trained Transformer (GPT-4) excel in numerical reasoning tasks i.e., temporal statistical analysis, achieve competitive correlation with human judgments on feature engineering tasks while struggle to some extent on domain specific knowledge reasoning, where domain-specific knowledge is required., Comment: 12 pages

Let ${\Bbb G}$ be a locally compact quantum group and ${\mathcal T}(L^2({\Bbb G}))$ be the Banach algebra of trace class operators on $L^2({\Bbb G})$ with the convolution induced by the right fundamental unitary of ${\Bbb G}$. We study the space of harmonic operators $\widetilde{\mathcal H}_\omega$ in ${\mathcal B}(L^2({\Bbb G}))$ associated to a contractive element $\omega\in {\mathcal T}(L^2({\Bbb G}))$. We characterize the existence of non-zero harmonic operators in ${\mathcal K}(L^2({\Bbb G}))$ and relate them with some properties of the quantum group ${\Bbb G}$, such as finiteness, amenability and co-amenability.

Assessing response quality to instructions in language models is vital but challenging due to the complexity of human language across different contexts. This complexity often results in ambiguous or inconsistent interpretations, making accurate assessment difficult. To address this issue, we propose a novel Uncertainty-aware Reward Model (URM) that introduces a robust uncertainty estimation for the quality of paired responses based on Bayesian approximation. Trained with preference datasets, our uncertainty-enabled proxy not only scores rewards for responses but also evaluates their inherent uncertainty. Empirical results demonstrate significant benefits of incorporating the proposed proxy into language model training. Our method boosts the instruction following capability of language models by refining data curation for training and improving policy optimization objectives, thereby surpassing existing methods by a large margin on benchmarks such as Vicuna and MT-bench. These findings highlight that our proposed approach substantially advances language model training and paves a new way of harnessing uncertainty within language models., Comment: Accepted to ICML 2024

Existing anti-malware software and reverse engineering toolkits struggle with stealthy sub-OS rootkits due to limitations of run-time kernel-level monitoring. A malicious kernel-level driver can bypass OS-level anti-virus mechanisms easily. Although static analysis of such malware is possible, obfuscation and packing techniques complicate offline analysis. Moreover, current dynamic analyzers suffer from virtualization performance overhead and create detectable traces that allow modern malware to evade them. To address these issues, we present \textit{The Reversing Machine} (TRM), a new hypervisor-based memory introspection design for reverse engineering, reconstructing memory offsets, and fingerprinting evasive and obfuscated user-level and kernel-level malware. TRM proposes two novel techniques that enable efficient and transparent analysis of evasive malware: hooking a binary using suspended process creation for hypervisor-based memory introspection, and leveraging Mode-Based Execution Control (MBEC) to detect user/kernel mode transitions and memory access patterns. Unlike existing malware detection environments, TRM can extract full memory traces in user and kernel spaces and hook the entire target memory map to reconstruct arrays, structures within the operating system, and possible rootkits. We perform TRM-assisted reverse engineering of kernel-level structures and show that it can speed up manual reverse engineering by 75\% on average. We obfuscate known malware with the latest packing tools and successfully perform similarity detection. Furthermore, we demonstrate a real-world attack by deploying a modified rootkit onto a driver that bypasses state-of-the-art security auditing tools. We show that TRM can detect each threat and that, out of 24 state-of-the-art AV solutions, only TRM can detect the most advanced threats.

Overlapping Community Search (OCS) identifies nodes that interact with multiple communities based on a specified query. Existing community search approaches fall into two categories: algorithm-based models and Machine Learning-based (ML) models. Despite the long-standing focus on this topic within the database domain, current solutions face two major limitations: 1) Both approaches fail to address personalized user requirements in OCS, consistently returning the same set of nodes for a given query regardless of user differences. 2) Existing ML-based CS models suffer from severe training efficiency issues. In this paper, we formally redefine the problem of OCS. By analyzing the gaps in both types of approaches, we then propose a general solution for OCS named Sparse Subspace Filter (SSF), which can extend any ML-based CS model to enable personalized search in overlapping structures. To overcome the efficiency issue in the current models, we introduce Simplified Multi-hop Attention Networks (SMN), a lightweight yet effective community search model with larger receptive fields. To the best of our knowledge, this is the first ML-based study of overlapping community search. Extensive experiments validate the superior performance of SMN within the SSF pipeline, achieving a 13.73% improvement in F1-Score and up to 3 orders of magnitude acceleration in model efficiency compared to state-of-the-art approaches.

The flux of ultra-high energy cosmic rays reaching Earth above the ankle energy (5 EeV) can be described as a mixture of nuclei injected by extragalactic sources with very hard spectra and a low rigidity cutoff. Extragalactic magnetic fields existing between the Earth and the closest sources can affect the observed CR spectrum by reducing the flux of low-rigidity particles reaching Earth. We perform a combined fit of the spectrum and distributions of depth of shower maximum measured with the Pierre Auger Observatory including the effect of this magnetic horizon in the propagation of UHECRs in the intergalactic space. We find that, within a specific range of the various experimental and phenomenological systematics, the magnetic horizon effect can be relevant for turbulent magnetic field strengths in the local neighbourhood of order $B_{\rm rms}\simeq (50-100)\,{\rm nG}\,(20\rm{Mpc}/{d_{\rm s})( 100\,\rm{kpc}/L_{\rm coh}})^{1/2}$, with $d_{\rm s}$ the typical intersource separation and $L_{\rm coh}$ the magnetic field coherence length. When this is the case, the inferred slope of the source spectrum becomes softer and can be closer to the expectations of diffusive shock acceleration, i.e., $\propto E^{-2}$. An additional cosmic-ray population with higher source density and softer spectra, presumably also extragalactic and dominating the cosmic-ray flux at EeV energies, is also required to reproduce the overall spectrum and composition results for all energies down to 0.6~EeV., Comment: Matches published version

Rehabilitation robotics continues to confront substantial challenges, particularly in achieving smooth, safe, and intuitive human-robot interactions for upper limb motor training. Many current systems depend on complex mechanical designs, direct physical contact, and multiple sensors, which not only elevate costs but also reduce accessibility. Additionally, delivering seamless weight compensation and precise motion tracking remains a highly complex undertaking. To overcome these obstacles, we have developed a novel magnetic-based actuation mechanism for end-effector robotic rehabilitation. This innovative approach enables smooth, non-contact force transmission, significantly enhancing patient safety and comfort during upper limb training. To ensure consistent performance, we integrated an Extended Kalman Filter (EKF) alongside a controller for real-time position tracking, allowing the system to maintain high accuracy or recover even in the event of sensor malfunction or failure. In a user study with 12 participants, 75% rated the system highly for its smoothness, while 66.7% commended its safety and effective weight compensation. The EKF demonstrated precise tracking performance, with root mean square error (RMSE) values remaining within acceptable limits (under 2 cm). By combining magnetic actuation with advanced closed-loop control algorithms, this system marks a significant advancement in the field of upper limb rehabilitation robotics., Comment: 7 pages, 9 figures, journal paper

Diffusion-weighted MRI (DWI) is essential for stroke diagnosis, treatment decisions, and prognosis. However, image and disease variability hinder the development of generalizable AI algorithms with clinical value. We address this gap by presenting a novel ensemble algorithm derived from the 2022 Ischemic Stroke Lesion Segmentation (ISLES) challenge. ISLES'22 provided 400 patient scans with ischemic stroke from various medical centers, facilitating the development of a wide range of cutting-edge segmentation algorithms by the research community. Through collaboration with leading teams, we combined top-performing algorithms into an ensemble model that overcomes the limitations of individual solutions. Our ensemble model achieved superior ischemic lesion detection and segmentation accuracy on our internal test set compared to individual algorithms. This accuracy generalized well across diverse image and disease variables. Furthermore, the model excelled in extracting clinical biomarkers. Notably, in a Turing-like test, neuroradiologists consistently preferred the algorithm's segmentations over manual expert efforts, highlighting increased comprehensiveness and precision. Validation using a real-world external dataset (N=1686) confirmed the model's generalizability. The algorithm's outputs also demonstrated strong correlations with clinical scores (admission NIHSS and 90-day mRS) on par with or exceeding expert-derived results, underlining its clinical relevance. This study offers two key findings. First, we present an ensemble algorithm (https://github.com/Tabrisrei/ISLES22_Ensemble) that detects and segments ischemic stroke lesions on DWI across diverse scenarios on par with expert (neuro)radiologists. Second, we show the potential for biomedical challenge outputs to extend beyond the challenge's initial objectives, demonstrating their real-world clinical applicability.

Building embodied AI systems that can follow arbitrary language instructions in any 3D environment is a key challenge for creating general AI. Accomplishing this goal requires learning to ground language in perception and embodied actions, in order to accomplish complex tasks. The Scalable, Instructable, Multiworld Agent (SIMA) project tackles this by training agents to follow free-form instructions across a diverse range of virtual 3D environments, including curated research environments as well as open-ended, commercial video games. Our goal is to develop an instructable agent that can accomplish anything a human can do in any simulated 3D environment. Our approach focuses on language-driven generality while imposing minimal assumptions. Our agents interact with environments in real-time using a generic, human-like interface: the inputs are image observations and language instructions and the outputs are keyboard-and-mouse actions. This general approach is challenging, but it allows agents to ground language across many visually complex and semantically rich environments while also allowing us to readily run agents in new environments. In this paper we describe our motivation and goal, the initial progress we have made, and promising preliminary results on several diverse research environments and a variety of commercial video games.

Few-shot Class-Incremental Learning (FSCIL) poses the challenge of retaining prior knowledge while learning from limited new data streams, all without overfitting. The rise of Vision-Language models (VLMs) has unlocked numerous applications, leveraging their existing knowledge to fine-tune on custom data. However, training the whole model is computationally prohibitive, and VLMs while being versatile in general domains still struggle with fine-grained datasets crucial for many applications. We tackle these challenges with two proposed simple modules. The first, Session-Specific Prompts (SSP), enhances the separability of image-text embeddings across sessions. The second, Hyperbolic distance, compresses representations of image-text pairs within the same class while expanding those from different classes, leading to better representations. Experimental results demonstrate an average 10-point increase compared to baselines while requiring at least 8 times fewer trainable parameters. This improvement is further underscored on our three newly introduced fine-grained datasets.

Model Predictive Control (MPC) has exhibited remarkable capabilities in optimizing objectives and meeting constraints. However, the substantial computational burden associated with solving the Optimal Control Problem (OCP) at each triggering instant introduces significant delays between state sampling and control application. These delays limit the practicality of MPC in resource-constrained systems when engaging in complex tasks. The intuition to address this issue in this paper is that by predicting the successor state, the controller can solve the OCP one time step ahead of time thus avoiding the delay of the next action. To this end, we compute deviations between real and nominal system states, predicting forthcoming real states as initial conditions for the imminent OCP solution. Anticipatory computation stores optimal control based on current nominal states, thus mitigating the delay effects. Additionally, we establish an upper bound for linearization error, effectively linearizing the nonlinear system, reducing OCP complexity, and enhancing response speed. We provide empirical validation through two numerical simulations and corresponding real-world robot tasks, demonstrating significant performance improvements and augmented response speed (up to $90\%$) resulting from the seamless integration of our proposed approach compared to conventional time-triggered MPC strategies., Comment: arXiv admin note: text overlap with arXiv:2103.09693

The Software Supply Chain (SSC) has captured considerable attention from attackers seeking to infiltrate systems and undermine organizations. There is evidence indicating that adversaries utilize Social Engineering (SocE) techniques specifically aimed at software developers. That is, they interact with developers at critical steps in the Software Development Life Cycle (SDLC), such as accessing Github repositories, incorporating code dependencies, and obtaining approval for Pull Requests (PR) to introduce malicious code. This paper aims to comprehensively explore the existing and emerging SocE tactics employed by adversaries to trick Software Engineers (SWEs) into delivering malicious software. By analyzing a diverse range of resources, which encompass established academic literature and real-world incidents, the paper systematically presents an overview of these manipulative strategies within the realm of the SSC. Such insights prove highly beneficial for threat modeling and security gap analysis., Comment: 7 pages, 2 figures

Neural networks successfully capture the computational power of the human brain for many tasks. Similarly inspired by the brain architecture, Nearest Neighbor (NN) representations is a novel approach of computation. We establish a firmer correspondence between NN representations and neural networks. Although it was known how to represent a single neuron using NN representations, there were no results even for small depth neural networks. Specifically, for depth-2 threshold circuits, we provide explicit constructions for their NN representation with an explicit bound on the number of bits to represent it. Example functions include NN representations of convex polytopes (AND of threshold gates), IP2, OR of threshold gates, and linear or exact decision lists., Comment: This paper is accepted to ISIT 2024. 2nd version has revisions for better clarity, more citations, and more explanation in the proofs. No results are changed

The Nearest Neighbor (NN) Representation is an emerging computational model that is inspired by the brain. We study the complexity of representing a neuron (threshold function) using the NN representations. It is known that two anchors (the points to which NN is computed) are sufficient for a NN representation of a threshold function, however, the resolution (the maximum number of bits required for the entries of an anchor) is $O(n\log{n})$. In this work, the trade-off between the number of anchors and the resolution of a NN representation of threshold functions is investigated. We prove that the well-known threshold functions EQUALITY, COMPARISON, and ODD-MAX-BIT, which require 2 or 3 anchors and resolution of $O(n)$, can be represented by polynomially large number of anchors in $n$ and $O(\log{n})$ resolution. We conjecture that for all threshold functions, there are NN representations with polynomially large size and logarithmic resolution in $n$., Comment: This paper is accepted to ISIT 2024. 2nd version had revisions for better clarity, fixing of typos. No results are changed

The recent success of Large Language Models (LLM) in a wide range of Natural Language Processing applications opens the path towards novel Question Answering Systems over Knowledge Graphs leveraging LLMs. However, one of the main obstacles preventing their implementation is the scarcity of training data for the task of translating questions into corresponding SPARQL queries, particularly in the case of domain-specific KGs. To overcome this challenge, in this study, we evaluate several strategies for fine-tuning the OpenLlama LLM for question answering over life science knowledge graphs. In particular, we propose an end-to-end data augmentation approach for extending a set of existing queries over a given knowledge graph towards a larger dataset of semantically enriched question-to-SPARQL query pairs, enabling fine-tuning even for datasets where these pairs are scarce. In this context, we also investigate the role of semantic "clues" in the queries, such as meaningful variable names and inline comments. Finally, we evaluate our approach over the real-world Bgee gene expression knowledge graph and we show that semantic clues can improve model performance by up to 33% compared to a baseline with random variable names and no comments included., Comment: To appear in Proceedings of SWAT4HCLS 2024: Semantic Web Tools and Applications for Healthcare and Life Sciences

There have been different reports of developing Brain-Computer Interface (BCI) platforms to investigate the noninvasive electroencephalography (EEG) signals associated with plan-to-grasp tasks in humans. However, these reports were unable to clearly show evidence of emerging neural activity from the planning (observation) phase - dominated by the vision cortices - to grasp execution - dominated by the motor cortices. In this study, we developed a novel vision-based grasping BCI platform that distinguishes different grip types (power and precision) through the phases of plan-to-grasp tasks using EEG signals. Using our platform and extracting features from Filter Bank Common Spatial Patterns (FBCSP), we show that frequency-band specific EEG contains discriminative spatial patterns present in both the observation and movement phases. Support Vector Machine (SVM) classification (power vs precision) yielded high accuracy percentages of 74% and 68% for the observation and movement phases in the alpha band, respectively., Comment: 4 pages, 4 figures, 1 table, conference paper

The bispectrum stands out as a revolutionary tool in frequency domain analysis, leaping the usual power spectrum by capturing crucial phase information between frequency components. In our innovative study, we have utilized the bispectrum to analyze and decode complex grasping movements, gathering EEG data from five human subjects. We put this data through its paces with three classifiers, focusing on both magnitude and phase-related features. The results highlight the bispectrum's incredible ability to delve into neural activity and differentiate between various grasping motions with the Support Vector Machine (SVM) classifier emerging as a standout performer. In binary classification, it achieved a remarkable 97\% accuracy in identifying power grasp, and in the more complex multiclass tasks, it maintained an impressive 94.93\% accuracy. This finding not only underscores the bispectrum's analytical strength but also showcases the SVM's exceptional capability in classification, opening new doors in our understanding of movement and neural dynamics., Comment: 4 pages, 3 figures, 12 references, 4 authors, conference paper

This study aims to enhance BCI applications for individuals with motor impairments by comparing the effectiveness of tripolar EEG (tEEG) with conventional EEG. The focus is on interpreting and decoding various grasping movements, such as power grasp and precision grasp. The goal is to determine which EEG technology is more effective in processing and translating grasp related neural signals. The approach involved experimenting on ten healthy participants who performed two distinct grasp movements: power grasp and precision grasp, with a no movement condition serving as the baseline. Our research presents a thorough comparison between EEG and tEEG in decoding grasping movements. This comparison spans several key parameters, including signal to noise ratio (SNR), spatial resolution via functional connectivity, ERPs, and wavelet time frequency analysis. Additionally, our study involved extracting and analyzing statistical features from the wavelet coefficients, and both binary and multiclass classification methods were employed. Four machine learning algorithms were used to evaluate the decoding accuracies. Our results indicated that tEEG demonstrated superior performance over conventional EEG in various aspects. This included a higher signal to noise ratio, enhanced spatial resolution, and more informative data in ERPs and wavelet time frequency analysis. The use of tEEG led to notable improvements in decoding accuracy for differentiating movement types. Specifically, tEEG achieved around 90% accuracy in binary and 75.97% for multiclass classification. These results are markedly better than those from standard EEG, which recorded a maximum of 77.85% and 61.27% in similar tasks, respectively. These findings highlight the superior effectiveness of tEEG over EEG in decoding grasp types and its competitive or superior performance in complex classifications compared with existing research.

This research aims to decode hand grasps from Electroencephalograms (EEGs) for dexterous neuroprosthetic development and Brain-Computer Interface (BCI) applications, especially for patients with motor disorders. Particularly, it focuses on distinguishing two complex natural power and precision grasps in addition to a neutral condition as a no-movement condition using a new EEG-based BCI platform and wavelet signal processing. Wavelet analysis involved generating time-frequency and topographic maps from wavelet power coefficients. Then, by using machine learning techniques with novel wavelet features, we achieved high average accuracies: 85.16% for multiclass, 95.37% for No-Movement vs Power, 95.40% for No-Movement vs Precision, and 88.07% for Power vs Precision, demonstrating the effectiveness of these features in EEG-based grasp differentiation. In contrast to previous studies, a critical part of our study was permutation feature importance analysis, which highlighted key features for grasp classification. It revealed that the most crucial brain activities during grasping occur in the motor cortex, within the alpha and beta frequency bands. These insights demonstrate the potential of wavelet features in real-time neuroprosthetic technology and BCI applications.

We study the problem of online conditional distribution estimation with \emph{unbounded} label sets under local differential privacy. Let $\mathcal{F}$ be a distribution-valued function class with unbounded label set. We aim at estimating an \emph{unknown} function $f\in \mathcal{F}$ in an online fashion so that at time $t$ when the context $\boldsymbol{x}_t$ is provided we can generate an estimate of $f(\boldsymbol{x}_t)$ under KL-divergence knowing only a privatized version of the true labels sampling from $f(\boldsymbol{x}_t)$. The ultimate objective is to minimize the cumulative KL-risk of a finite horizon $T$. We show that under $(\epsilon,0)$-local differential privacy of the privatized labels, the KL-risk grows as $\tilde{\Theta}(\frac{1}{\epsilon}\sqrt{KT})$ upto poly-logarithmic factors where $K=|\mathcal{F}|$. This is in stark contrast to the $\tilde{\Theta}(\sqrt{T\log K})$ bound demonstrated by Wu et al. (2023a) for bounded label sets. As a byproduct, our results recover a nearly tight upper bound for the hypothesis selection problem of gopi et al. (2020) established only for the batch setting.

We study the problem of oracle-efficient hybrid online learning when the features are generated by an unknown i.i.d. process and the labels are generated adversarially. Assuming access to an (offline) ERM oracle, we show that there exists a computationally efficient online predictor that achieves a regret upper bounded by $\tilde{O}(T^{\frac{3}{4}})$ for a finite-VC class, and upper bounded by $\tilde{O}(T^{\frac{p+1}{p+2}})$ for a class with $\alpha$ fat-shattering dimension $\alpha^{-p}$. This provides the first known oracle-efficient sublinear regret bounds for hybrid online learning with an unknown feature generation process. In particular, it confirms a conjecture of Lazaric and Munos (JCSS 2012). We then extend our result to the scenario of shifting distributions with $K$ changes, yielding a regret of order $\tilde{O}(T^{\frac{4}{5}}K^{\frac{1}{5}})$. Finally, we establish a regret of $\tilde{O}((K^{\frac{2}{3}}(\log|\mathcal{H}|)^{\frac{1}{3}}+K)\cdot T^{\frac{4}{5}})$ for the contextual $K$-armed bandits with a finite policy set $\mathcal{H}$, i.i.d. generated contexts from an unknown distribution, and adversarially generated costs.

We test the predictions of hadronic interaction models regarding the depth of maximum of air-shower profiles, $X_{max}$, and ground-particle signals in water-Cherenkov detectors at 1000 m from the shower core, $S(1000)$, using the data from the fluorescence and surface detectors of the Pierre Auger Observatory. The test consists in fitting the measured two-dimensional ($S(1000)$, $X_{max}$) distributions using templates for simulated air showers produced with hadronic interaction models EPOS-LHC, QGSJet II-04, Sibyll 2.3d and leaving the scales of predicted $X_{max}$ and the signals from hadronic component at ground as free fit parameters. The method relies on the assumption that the mass composition remains the same at all zenith angles, while the longitudinal shower development and attenuation of ground signal depend on the mass composition in a correlated way. The analysis was applied to 2239 events detected by both the fluorescence and surface detectors of the Pierre Auger Observatory with energies between $10^{18.5}$ to $10^{19.0}$ eV and zenith angles below $60^\circ$. We found, that within the assumptions of the method, the best description of the data is achieved if the predictions of the hadronic interaction models are shifted to deeper $X_{max}$ values and larger hadronic signals at all zenith angles. Given the magnitude of the shifts and the data sample size, the statistical significance of the improvement of data description using the modifications considered in the paper is larger than $5\sigma$ even for any linear combination of experimental systematic uncertainties., Comment: Published version