Search

In this paper, we shall present an interesting and significant refinement of a classical result of Cauchy about the moduli of the zeros of a quaternionic polynomial. As an application of this result we shall obtain zero-free regions of polynomials having quaternionic coefficients., Comment: 8 pages, no figures

A systematic analysis of $E2$ matrix elements of $^{72}$Ge, $^{76}$Ge, $^{168}$Er, $^{186}$Os, $^{188}$Os, $^{190}$Os, $^{192}$Os and $^{194}$Pt nuclides is performed using the beyond mean-field approach of triaxial projected shell model (TPSM). For these nuclei, large sets of $E2$ matrix elements have been deduced from the multi-step Coulomb excitation experiments, and it is shown that TPSM approach provides a reasonable description of the measured transitions. We have evaluated 1496 $E2$ matrix elements up to spin, $I=10$ for the eight nuclei studied, and tabulate them for future experimental and theoretical comparisons. Further, shape invariant analysis has been performed with the calculated $E2$ transitions using the Kumar-Cline sum rules. It is inferred from the analysis that the resulting shape, after configuration mixing of the quasiparticle states, transforms from $\gamma$-rigid to that of $\gamma$-soft for some nuclei, in conformity with the experimental data., Comment: 24 pages, 14 figures

This paper argues that generating output tokens is more effective than using pooled representations for prediction tasks because token-level generation retains more mutual information. Since LLMs are trained on massive text corpora using next-token prediction, generation aligns naturally with their learned behavior. Using the Data Processing Inequality (DPI), we provide both theoretical and empirical evidence supporting this claim. However, autoregressive models face two key challenges when used for prediction: (1) exposure bias, where the model sees ground truth tokens during training but relies on its own predictions during inference, leading to errors, and (2) format mismatch, where discrete tokens do not always align with the tasks required output structure. To address these challenges, we introduce PredGen(Predicting Through Generating), an end to end framework that (i) uses scheduled sampling to reduce exposure bias, and (ii) introduces a task adapter to convert the generated tokens into structured outputs. Additionally, we introduce Writer-Director Alignment Loss (WDAL), which ensures consistency between token generation and final task predictions, improving both text coherence and numerical accuracy. We evaluate PredGen on multiple classification and regression benchmarks. Our results show that PredGen consistently outperforms standard baselines, demonstrating its effectiveness in structured prediction tasks., Comment: Preprint paper

Drug-target interactions are critical for understanding biological processes and advancing drug discovery. However, traditional methods such as ComplEx-SE, TransE, and DistMult struggle with unseen relationships and negative triplets, which limits their effectiveness in drug-target prediction. To address these challenges, we propose Multi-Context-Aware Sampling (MuCoS), an efficient and positively accurate method for drug-target prediction. MuCoS reduces computational complexity by prioritizing neighbors of higher density to capture informative structural patterns. These optimized neighborhood representations are integrated with BERT, enabling contextualized embeddings for accurate prediction of missing relationships or tail entities. MuCoS avoids the need for negative triplet sampling, reducing computation while improving performance over unseen entities and relations. Experiments on the KEGG50k biomedical dataset show that MuCoS improved over existing models by 13\% on MRR, 7\% on Hits@1, 4\% on Hits@3, and 18\% on Hits@10 for the general relationship, and by 6\% on MRR, 1\% on Hits@1, 3\% on Hits@3, and 12\% on Hits@10 for prediction of drug-target relationship.

In this paper, we are concerned with the problem of locating the zeros of polynomials of a quaternionic variable with quaternionic coefficients. We derive some new Cauchy bounds for the zeros of a polynomial by virtue of maximum modulus theorem. Our results will generalise some recently proved results about the distribution of zeros of a quaternionic polynomial., Comment: 6,0

While large models pre-trained on high-quality data exhibit excellent performance across various reasoning tasks, including mathematical reasoning (e.g. GSM8k, MultiArith), specializing smaller models to excel at mathematical reasoning remains a challenging problem. Common approaches to address this challenge include knowledge distillation, where smaller student models learn from large pre-trained teacher models, and data augmentation, such as rephrasing questions. Despite these efforts, smaller models struggle with arithmetic computations, leading to errors in mathematical reasoning. In this work, we focus on leveraging a programmatically generated arithmetic dataset to enhance the reasoning capabilities of smaller models. We investigate two key approaches to incorporate this dataset -- (1) intermediate fine-tuning, where a model is fine-tuned on the arithmetic dataset before being trained on a reasoning dataset, and (2) integrating the arithmetic dataset into the instruction-tuning mixture, allowing the model to learn arithmetic skills alongside general instruction-following abilities. Our experiments on multiple reasoning benchmarks demonstrate that incorporating an arithmetic dataset, whether through targeted fine-tuning or within the instruction-tuning mixture, enhances the models' arithmetic capabilities, which in turn improves their mathematical reasoning performance., Comment: Preprint

We study unitary orthonormal bases in the sense of Pimsner and Popa for inclusions $(\mathcal{B}\subseteq \mathcal{A}, E),$ where $\mathcal{A}, \mathcal{B}$ are finite dimensional von Neumann algebras and $E$ is a conditional expectation map from $\mathcal{A}$ onto $\mathcal{B}$. It is shown that existence of such bases requires that the associated inclusion matrix satisfies a spectral condition forcing dimension vectors to be Perron-Frobenius eigenvectors and the conditional expectation map preserves the Markov trace. Subject to these conditions, explicit unitary orthonormal bases are constructed if either one of the algebras is abelian or simple. They generalize complex Hadamard matrices, Weyl unitary bases, and a recent work of Crann et al which correspond to the special cases of $\mathcal{A}$ being abelian, simple, and general multi-matrix algebras respectively with $\mathcal{B}$ being the algebra of complex numbers. For the first time $\mathcal{B}$ is more general. As an application of these results it is shown that if $(\mathcal{B}\subseteq \mathcal{A}, E),$ admits a unitary orthonormal basis then the Connes-St{\o}rmer relative entropy $H(\mathcal{A}_1|\mathcal{A})$ equals the logarithm of the square of the norm of the inclusion matrix, where $\mathcal{A}_1$ denotes the Jones basic construction of the inclusion. As a further application, we prove the existence of unitary orthonormal bases for a large class of depth 2 subfactors with abelian relative commutant., Comment: 23 pages, no figures

We present MicroBooNE's first search for dark sector $e^+e^-$ explanations of the long-standing MiniBooNE anomaly. The MiniBooNE anomaly has garnered significant attention over the past 20 years including previous MicroBooNE investigations into both anomalous electron and photon excesses, but its origin still remains unclear. In this letter, we provide the first direct test of dark sector models in which dark neutrinos, produced through neutrino-induced scattering, decay into missing energy and visible $e^+e^-$ pairs comprising the MiniBooNE anomaly. Many such models have recently gained traction as a viable solution to the anomaly while evading past bounds. Using an exposure of $6.87 \times 10^{20}$ protons-on-target in the Booster Neutrino Beam, we implement a selection targeting forward-going, coherently produced $e^+e^-$ events. After unblinding, we observe 95 events, which we compare with the constrained background-only prediction of $69.7 \pm 17.3$. This analysis sets the world's first direct limits on these dark sector models and, at the 95\% confidence level, excludes the majority of the parameter space viable as a solution to the MiniBooNE anomaly., Comment: 7 pages, 5 figures, Supplemental Materials included in Ancillary files

User profile modeling plays a key role in personalized systems, as it requires building accurate profiles and updating them with new information. In this paper, we present two high-quality open-source user profile datasets: one for profile construction and another for profile updating. These datasets offer a strong basis for evaluating user profile modeling techniques in dynamic settings. We also show a methodology that uses large language models (LLMs) to tackle both profile construction and updating. Our method uses a probabilistic framework to predict user profiles from input text, allowing for precise and context-aware profile generation. Our experiments demonstrate that models like Mistral-7b and Llama2-7b perform strongly in both tasks. LLMs improve the precision and recall of the generated profiles, and high evaluation scores confirm the effectiveness of our approach.

Artificial Intelligence (AI) systems are increasingly intertwined with daily life, assisting users in executing various tasks and providing guidance on decision-making. This integration introduces risks of AI-driven manipulation, where such systems may exploit users' cognitive biases and emotional vulnerabilities to steer them toward harmful outcomes. Through a randomized controlled trial with 233 participants, we examined human susceptibility to such manipulation in financial (e.g., purchases) and emotional (e.g., conflict resolution) decision-making contexts. Participants interacted with one of three AI agents: a neutral agent (NA) optimizing for user benefit without explicit influence, a manipulative agent (MA) designed to covertly influence beliefs and behaviors, or a strategy-enhanced manipulative agent (SEMA) employing explicit psychological tactics to reach its hidden objectives. By analyzing participants' decision patterns and shifts in their preference ratings post-interaction, we found significant susceptibility to AI-driven manipulation. Particularly, across both decision-making domains, participants interacting with the manipulative agents shifted toward harmful options at substantially higher rates (financial, MA: 62.3%, SEMA: 59.6%; emotional, MA: 42.3%, SEMA: 41.5%) compared to the NA group (financial, 35.8%; emotional, 12.8%). Notably, our findings reveal that even subtle manipulative objectives (MA) can be as effective as employing explicit psychological strategies (SEMA) in swaying human decision-making. By revealing the potential for covert AI influence, this study highlights a critical vulnerability in human-AI interactions, emphasizing the need for ethical safeguards and regulatory frameworks to ensure responsible deployment of AI technologies and protect human autonomy., Comment: Work in progress

The Pandora Software Development Kit and algorithm libraries perform reconstruction of neutrino interactions in liquid argon time projection chamber detectors. Pandora is the primary event reconstruction software used at the Deep Underground Neutrino Experiment, which will operate four large-scale liquid argon time projection chambers at the far detector site in South Dakota, producing high-resolution images of charged particles emerging from neutrino interactions. While these high-resolution images provide excellent opportunities for physics, the complex topologies require sophisticated pattern recognition capabilities to interpret signals from the detectors as physically meaningful objects that form the inputs to physics analyses. A critical component is the identification of the neutrino interaction vertex. Subsequent reconstruction algorithms use this location to identify the individual primary particles and ensure they each result in a separate reconstructed particle. A new vertex-finding procedure described in this article integrates a U-ResNet neural network performing hit-level classification into the multi-algorithm approach used by Pandora to identify the neutrino interaction vertex. The machine learning solution is seamlessly integrated into a chain of pattern-recognition algorithms. The technique substantially outperforms the previous BDT-based solution, with a more than 20\% increase in the efficiency of sub-1\,cm vertex reconstruction across all neutrino flavours., Comment: 32 pages, 18 figures

This article presents the first search for neutrino-induced neutral current coherent single-photon production (NC coherent 1$\gamma$). The search makes use of data from the MicroBooNE 85-tonne active volume liquid argon time projection chamber detector, situated in the Fermilab Booster Neutrino Beam (BNB), with an average neutrino energy of $\langle E_{\nu}\rangle \sim 0.8$ GeV. A targeted selection of candidate neutrino interactions with a single photon-like electromagnetic shower in the final state and no visible vertex activity was developed to search for the NC coherent 1$\gamma$ process, along with two auxiliary selections used to constrain the dominant background from NC$\pi^0$ production. With an integrated exposure of $6.87 \times 10^{20}$ protons on target delivered by the BNB, we set the world's first limit for this rare process, corresponding to an upper limit on the flux-averaged cross section of $\sigma<1.49 \times 10^{-41}\text{cm}^2$ at 90\% C.L., Comment: 20 pages, 17 figures

We present an inclusive search for anomalous production of single-photon events from neutrino interactions in the MicroBooNE experiment. The search and its signal definition are motivated by the previous observation of a low-energy excess of electromagnetic shower events from the MiniBooNE experiment. We use the Wire-Cell reconstruction framework to select a sample of inclusive single-photon final-state interactions with a final efficiency and purity of 7.0% and 40.2%, respectively. We leverage simultaneous measurements of sidebands of charged current $\nu_{\mu}$ interactions and neutral current interactions producing $\pi^{0}$ mesons to constrain signal and background predictions and reduce uncertainties. We perform a blind analysis using a dataset collected from February 2016 to July 2018, corresponding to an exposure of $6.34\times10^{20}$ protons on target from the Booster Neutrino Beam (BNB) at Fermilab. In the full signal region, we observe agreement between the data and the prediction, with a goodness-of-fit $p$-value of 0.11. We then isolate a sub-sample of these events containing no visible protons, and observe $93\pm22\text{(stat.)}\pm35\text{(syst.)}$ data events above prediction, corresponding to just above $2\sigma$ local significance, concentrated at shower energies below 600 MeV., Comment: 9 pages, 6 figures

We report results from an updated search for neutral current (NC) resonant $\Delta$(1232) baryon production and subsequent $\Delta$ radiative decay (NC $\Delta\rightarrow N \gamma$). We consider events with and without final state protons; events with a proton can be compared with the kinematics of a $\Delta(1232)$ baryon decay, while events without a visible proton represent a more generic phase space. In order to maximize sensitivity to each topology, we simultaneously make use of two different reconstruction paradigms, Pandora and Wire-Cell, which have complementary strengths, and select mostly orthogonal sets of events. Considering an overall scaling of the NC $\Delta\rightarrow N \gamma$ rate as an explanation of the MiniBooNE anomaly, our data exclude this hypothesis at 94.4% CL. When we decouple the expected correlations between NC $\Delta\rightarrow N \gamma$ events with and without final state protons, and allow independent scaling of both types of events, our data exclude explanations in which excess events have associated protons, and do not exclude explanations in which excess events have no associated protons.

Geometrical complexities in natural fault zones, such as steps and gaps, pose a challenge in seismic hazard studies as they can act as barriers to seismic ruptures. In this study, we propose a criterion, which is based on the rate-and-state equation, to estimate the efficiency of an earthquake rupture to jump two spatially disconnected faults. The proposed jump criterion is tested using a 2D quasi-dynamic numerical simulations of the seismic cycle. The criterion successfully predicts fault jumps where the coulomb stress change fails to do so. The criterion includes the coulomb stress change as a parameter but is also dependent on other important parameters among which the absolute normal stress on the fault which the rupture is to jump to. Based on the criterion, the maximum jump distance increases with decreasing absolute normal stress, i.e. as the rupture process occurs closer to the surface or as pore pressure increases. The criterion implies that an earthquake can jump to an infinite distance at the surface if the normal stress is allowed to go to zero. Thus, the properties of the surface layers are of the outmost importance in terms of maximum rupture jump distance allowed. The absolute normal stress is the main controlling parameter followed by the uncertainty on the slip of an earthquake, which controls the coulomb stress impact on the receiver fault. Finally, we also propose a method to compute probabilities of earthquakes rupture to jump, which allows to consider uncertainties in geometrical configurations between two faults.

Large language models (LLMs) have shown great potential in natural language processing tasks, but their application to machine translation (MT) remains challenging due to pretraining on English-centric data and the complexity of reinforcement learning from human feedback (RLHF). Direct Preference Optimization (DPO) has emerged as a simpler and more efficient alternative, but its performance depends heavily on the quality of preference data. To address this, we propose Confidence-Reward driven Preference Optimization (CRPO), a novel method that combines reward scores with model confidence to improve data selection for fine-tuning. CRPO selects challenging sentence pairs where the model is uncertain or underperforms, leading to more effective learning. While primarily designed for LLMs, CRPO also generalizes to encoder-decoder models like NLLB, demonstrating its versatility. Empirical results show that CRPO outperforms existing methods such as RS-DPO, RSO and MBR score in both translation accuracy and data efficiency.

Solving for time evolution of a many particle system whose dynamics is governed by Lindblad equation is hard. We extend the use of transfer matrix approach to a class of Linblad equations that admit a closed hierarchy of two point correlators. An example that we treat is the XX spin chain, i.e., free fermions, subject to the local on-site dephasing, but can be extended to other Hermitian dissipators, e.g., non-local dephasing. We find a simple expression of the Green's function in the Laplace domain. The method can be used to get analytical results in the thermodynamic limit, for instance, to get the evolution of magnetization density and to explicitly see the cross over between the ballistic and diffusive behavior, or to show that the correlations between operators at distance $l$ decay with time as $1/t^{\lceil l/2 \rceil+1/2}$. It also provides a fast numerical method to determine the evolution of the density with a complexity scaling with the system size more favorably than in previous methods, easily allowing one to study systems with $\sim 10^6$ spins., Comment: 11 pages and 4 figures

Here we present experimental evidence of the pseudo-doublet ground state of the non-Kramers compound SrTm2O4, based on specific heat, magnetic entropy and electron paramagnetic resonance. We demonstrate that the two crystallographic Tm3+ sites give rise to distinct single-ion anisotropies, and by extension, SrTm2O4 hosts two magnetic sublattices. Inelastic neutron scattering reveals low-lying dispersing crystal-field excitations, which we modelled using an effective charge model and mean field random phase approximation. The extracted magnetic exchange interactions are both antiferromagnetic and frustrated for both chains. Interchain magnetic exchange interactions are negligible. The strength of the magnetic exchange interactions in relation to the size of crystal field gaps, together with the frustration and low dimensionality, force the system to remain paramagnetic down to the lowest experimentally reachable temperature despite the pseudo-doublet nature of its ground state.

Acquiring and annotating surgical data is often resource-intensive, ethical constraining, and requiring significant expert involvement. While generative AI models like text-to-image can alleviate data scarcity, incorporating spatial annotations, such as segmentation masks, is crucial for precision-driven surgical applications, simulation, and education. This study introduces both a novel task and method, SimGen, for Simultaneous Image and Mask Generation. SimGen is a diffusion model based on the DDPM framework and Residual U-Net, designed to jointly generate high-fidelity surgical images and their corresponding segmentation masks. The model leverages cross-correlation priors to capture dependencies between continuous image and discrete mask distributions. Additionally, a Canonical Fibonacci Lattice (CFL) is employed to enhance class separability and uniformity in the RGB space of the masks. SimGen delivers high-fidelity images and accurate segmentation masks, outperforming baselines across six public datasets assessed on image and semantic inception distance metrics. Ablation study shows that the CFL improves mask quality and spatial separation. Downstream experiments suggest generated image-mask pairs are usable if regulations limit human data release for research. This work offers a cost-effective solution for generating paired surgical images and complex labels, advancing surgical AI development by reducing the need for expensive manual annotations., Comment: 12 pages, 17 figures, 4 tables, project page at https://camma-public.github.io/endogen/

Target Speech Extraction (TSE) traditionally relies on explicit clues about the speaker's identity like enrollment audio, face images, or videos, which may not always be available. In this paper, we propose a text-guided TSE model StyleTSE that uses natural language descriptions of speaking style in addition to the audio clue to extract the desired speech from a given mixture. Our model integrates a speech separation network adapted from SepFormer with a bi-modality clue network that flexibly processes both audio and text clues. To train and evaluate our model, we introduce a new dataset TextrolMix with speech mixtures and natural language descriptions. Experimental results demonstrate that our method effectively separates speech based not only on who is speaking, but also on how they are speaking, enhancing TSE in scenarios where traditional audio clues are absent. Demos are at: https://mingyue66.github.io/TextrolMix/demo/, Comment: Accepted by ICASSP 2025

Long-period radio transients are a novel class of astronomical objects characterised by prolonged periods ranging from 18 minutes to 54 minutes. They exhibit highly polarised, coherent, beamed radio emission lasting only 10--100 seconds. The intrinsic nature of these objects is subject to speculation, with highly magnetised white dwarfs and neutron stars being the prevailing candidates. Here we present ASKAP J183950.5-075635.0 (hereafter, ASKAP J1839-0756), boasting the longest known period of this class at 6.45 hours. It exhibits emission characteristics of an ordered dipolar magnetic field, with pulsar-like bright main pulses and weaker interpulses offset by about half a period are indicative of an oblique or orthogonal rotator. This phenomenon, observed for the first time in a long-period radio transient, confirms that the radio emission originates from both magnetic poles and that the observed period corresponds to the rotation period. The spectroscopic and polarimetric properties of ASKAP J1839-0756 are consistent with a neutron star origin, and this object is a crucial piece of evidence in our understanding of long-period radio sources and their links to neutron stars., Comment: 44 pages, 14 figures, 2 tables This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this article is published in Nature Astronomy, and is available online at https://doi.org/10.1038/s41550-024-02452-z

Extended Reality (XR) enables a plethora of novel interactive shared experiences. Ideally, users are allowed to roam around freely, while audiovisual content is delivered wirelessly to their Head-Mounted Displays (HMDs). Therefore, truly immersive experiences will require massive amounts of data, in the range of tens of gigabits per second, to be delivered reliably at extremely low latencies. We identify Millimeter-Wave (mmWave) communications, at frequencies between 24 and 300 GHz, as a key enabler for such experiences. In this article, we show how the mmWave state of the art does not yet achieve sufficient performance, and identify several key active research directions expected to eventually pave the way for extremely-high-quality mmWave-enabled interactive multi-user XR., Comment: Published in IEEE Communications Magazine

We present the strongest limits to date on the mixing angle, $\theta$, with which a new scalar particle, $S$, mixes with the Higgs field in the mass range $100$ $MeV

While current high-resolution depth estimation methods achieve strong results, they often suffer from computational inefficiencies due to reliance on heavyweight models and multiple inference steps, increasing inference time. To address this, we introduce PatchRefiner V2 (PRV2), which replaces heavy refiner models with lightweight encoders. This reduces model size and inference time but introduces noisy features. To overcome this, we propose a Coarse-to-Fine (C2F) module with a Guided Denoising Unit for refining and denoising the refiner features and a Noisy Pretraining strategy to pretrain the refiner branch to fully exploit the potential of the lightweight refiner branch. Additionally, we introduce a Scale-and-Shift Invariant Gradient Matching (SSIGM) loss to enhance synthetic-to-real domain transfer. PRV2 outperforms state-of-the-art depth estimation methods on UnrealStereo4K in both accuracy and speed, using fewer parameters and faster inference. It also shows improved depth boundary delineation on real-world datasets like CityScape, ScanNet++, and KITTI, demonstrating its versatility across domains.

Forecasting is a task that is difficult to evaluate: the ground truth can only be known in the future. Recent work showing LLM forecasters rapidly approaching human-level performance begs the question: how can we benchmark and evaluate these forecasters instantaneously? Following the consistency check framework, we measure the performance of forecasters in terms of the consistency of their predictions on different logically-related questions. We propose a new, general consistency metric based on arbitrage: for example, if a forecasting AI illogically predicts that both the Democratic and Republican parties have 60% probability of winning the 2024 US presidential election, an arbitrageur can trade against the forecaster's predictions and make a profit. We build an automated evaluation system that generates a set of base questions, instantiates consistency checks from these questions, elicits the predictions of the forecaster, and measures the consistency of the predictions. We then build a standard, proper-scoring-rule forecasting benchmark, and show that our (instantaneous) consistency metrics correlate with LLM forecasters' ground truth Brier scores (which are only known in the future). We also release a consistency benchmark that resolves in 2028, providing a long-term evaluation tool for forecasting., Comment: 55 pages, 25 figures. Submitted to ICLR 2025

We study the homeomorphism groups of ordinals equipped with their order topology, focusing on successor ordinals whose limit capacity is also a successor. This is a rich family of groups that has connections to both permutation groups and homeomorphism groups of manifolds. For ordinals of Cantor--Bendixson degree one, we prove that the homeomorphism group is strongly distorted and uniformly perfect, and we classify its normal generators. As a corollary, we recover and provide a new proof of the classical result that the subgroup of finite permutations in the symmetric group on a countably infinite set is the maximal proper normal subgroup. For ordinals of higher Cantor--Bendixson degree, we establish a semi-direct product decomposition of the (pure) homeomorphism group. When the limit capacity is one, we further compute the abelianizations and determine normal generating sets of minimal cardinality for these groups., Comment: 28 pages

Learning-based solutions for long-tailed recognition face difficulties in generalizing on balanced test datasets. Due to imbalanced data prior, the learned \textit{a posteriori} distribution is biased toward the most frequent (head) classes, leading to an inferior performance on the least frequent (tail) classes. In general, the performance can be improved by removing such a bias by eliminating the effect of imbalanced prior modeled using the number of class samples (frequencies). We first observe that the \textit{effective prior} on the classes, learned by the model at the end of the training, can differ from the empirical prior obtained using class frequencies. Thus, we propose a novel approach to accurately model the effective prior of a trained model using \textit{a posteriori} probabilities. We propose to correct the imbalanced prior by adjusting the predicted \textit{a posteriori} probabilities (Prior2Posterior: P2P) using the calculated prior in a post-hoc manner after the training, and show that it can result in improved model performance. We present theoretical analysis showing the optimality of our approach for models trained with naive cross-entropy loss as well as logit adjusted loss. Our experiments show that the proposed approach achieves new state-of-the-art (SOTA) on several benchmark datasets from the long-tail literature in the category of logit adjustment methods. Further, the proposed approach can be used to inspect any existing method to capture the \textit{effective prior} and remove any residual bias to improve its performance, post-hoc, without model retraining. We also show that by using the proposed post-hoc approach, the performance of many existing methods can be improved further., Comment: IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2025

This Letter presents an investigation of low-energy electron-neutrino interactions in the Fermilab Booster Neutrino Beam by the MicroBooNE experiment, motivated by the excess of electron-neutrino-like events observed by the MiniBooNE experiment. This is the first measurement to use data from all five years of operation of the MicroBooNE experiment, corresponding to an exposure of $1.11\times 10^{21}$ protons on target, a $70\%$ increase on past results. Two samples of electron neutrino interactions without visible pions are used, one with visible protons and one without any visible protons. MicroBooNE data is compared to two empirical models that modify the predicted rate of electron-neutrino interactions in different variables in the simulation to match the unfolded MiniBooNE low energy excess. In the first model, this unfolding is performed as a function of electron neutrino energy, while the second model aims to match the observed shower energy and angle distributions of the MiniBooNE excess. This measurement excludes an electron-like interpretation of the MiniBooNE excess based on these models at $> 99\%$ CL$_\mathrm{s}$ in all kinematic variables., Comment: 8 pages, 5 figures, 1 table

Knowledge graphs (KGs) are valuable for representing structured, interconnected information across domains, enabling tasks like semantic search, recommendation systems and inference. A pertinent challenge with KGs, however, is that many entities (i.e., heads, tails) or relationships are unknown. Knowledge Graph Completion (KGC) addresses this by predicting these missing nodes or links, enhancing the graph's informational depth and utility. Traditional methods like TransE and ComplEx predict tail entities but struggle with unseen entities. Textual-based models leverage additional semantics but come with high computational costs, semantic inconsistencies, and data imbalance issues. Recent LLM-based models show improvement but overlook contextual information and rely heavily on entity descriptions. In this study, we introduce a contextualized BERT model for KGC that overcomes these limitations by utilizing the contextual information from neighbouring entities and relationships to predict tail entities. Our model eliminates the need for entity descriptions and negative triplet sampling, reducing computational demands while improving performance. Our model outperforms state-of-the-art methods on standard datasets, improving Hit@1 by 5.3% and 4.88% on FB15k-237 and WN18RR respectively, setting a new benchmark in KGC., Comment: MuslML Workshop, 38th Conference on Neural Information Processing Systems (NeurIPS 2024)

For many, it can be intimidating or even impossible to seek professional medical help if they have symptoms of an illness. As such, some people approach platforms like Reddit or Quora for a community-based conversation in an attempt to diagnose themselves. In this paper, we unearth what motivates people to share personal health information on these platforms. From an online survey and in-depth interviews, we present who this population of users are, and what, where, and why they are posting. Our evaluation finds that tech-savvy young adults are more likely to post on online platforms about potentially sensitive or highly specific topics for convenience, fast response, and a sense of community. Most importantly, we found that anonymity, distrust of physicians, and prior experience with platforms were key factors that affected behavior., Comment: 5 pages, 2 figures, Spring 2021 HCI Seminar, Columbia University, NY

Magnetic reconnection, a fundamental plasma process, is pivotal in understanding energy conversion and particle acceleration in astrophysical systems. While extensively studied in two-dimensional (2D) configurations, the dynamics of reconnection in three-dimensional (3D) systems remain under-explored. In this work, we extend the classical tearing mode instability to 3D by introducing a modulation along the otherwise uniform direction in a 2D equilibrium, given by $g(y)$, mimicking a flux tube-like configuration. We perform linear stability analysis (both analytically and numerically) and direct numerical simulations to investigate the effects of three-dimensionality. Our findings reveal that the 3D tearing instability exhibits reduced growth rates compared to 2D by a factor of $\int g(y)^{1/2} dy~/\int dy$, with the dispersion relation maintaining similar scaling characteristics. We show that the modulation introduces spatially varying resistive layer properties, which influence the reconnection dynamics. Remarkably, we find that Sweet-Parker scaling for the reconnection rate persists even in the absence of a guide field., Comment: 18 pages, 11 figures, comments are welcome

Audio Description (AD) plays a pivotal role as an application system aimed at guaranteeing accessibility in multimedia content, which provides additional narrations at suitable intervals to describe visual elements, catering specifically to the needs of visually impaired audiences. In this paper, we introduce $\mathrm{CA^3D}$, the pioneering unified Context-Aware Automatic Audio Description system that provides AD event scripts with precise locations in the long cinematic content. Specifically, $\mathrm{CA^3D}$ system consists of: 1) a Temporal Feature Enhancement Module to efficiently capture longer term dependencies, 2) an anchor-based AD event detector with feature suppression module that localizes the AD events and extracts discriminative feature for AD generation, and 3) a self-refinement module that leverages the generated output to tweak AD event boundaries from coarse to fine. Unlike conventional methods which rely on metadata and ground truth AD timestamp for AD detection and generation tasks, the proposed $\mathrm{CA^3D}$ is the first end-to-end trainable system that only uses visual cue. Extensive experiments demonstrate that the proposed $\mathrm{CA^3D}$ improves existing architectures for both AD event detection and script generation metrics, establishing the new state-of-the-art performances in the AD automation., Comment: 10 pages

We study an inflation model with nonminimal derivative coupling that features a coupling between the derivative of the inflaton field and the Einstein tensor. This model naturally amplifies curvature perturbations at small scales via gravitationally enhanced friction, a mechanism critical for the formation of primordial black holes and the associated production of potentially detectable scalar-induced gravitational waves. We derive analytical expressions for the primordial power spectrum, enabling efficient exploration of the model parameter space without requiring computationally intensive numerical solutions of the Mukhanov-Sasaki equation. Using the third data release of the Parkes Pulsar Timing Array (PPTA DR3), we constrain the model parameters characterizing the coupling function: $\phi_c = 3.7^{+0.3}_{-0.5} M_\mathrm{P}$, $\log_{10} \omega_L = 7.1^{+0.6}_{-0.3}$, and $\log_{10} \sigma = -8.3^{+0.3}_{-0.6}$ at 90\% confidence level. Our results demonstrate the growing capability of pulsar timing arrays to probe early Universe physics, complementing traditional cosmic microwave background observations by providing unique constraints on inflationary dynamics at small scales., Comment: 14 pages, 5figures. Accepted for publication as a Letter in Physical Review D

High-dimensional quantum entanglement in photons offers notable technological advancements over traditional qubit-based systems, including increased information density and enhanced security. However, such high-dimensional states are vulnerable to disruption by complex disordered media, presenting significant challenges in practical applications. Spatially-entangled photons are conventionally generated using a nonlinear crystal via spontaneous parametric down conversion (SPDC). While the effect of disorder on spatially entangled photons in the near field of the crystal is well understood, the impact of disorder in the far field is more complex. In this work, we present a systematic study of the randomization of two-photon correlations caused by arbitrary phase distortions in the far field by breaking it down into odd and even parity components. First, we theoretically show that the two-photon field is only sensitive to the even-parity part of the phase distortion. In follow-up experiments, we employ a deformable mirror to implement random phase distortions, separating the contributions of odd and even parity phases using Zernike polynomials. The experimental results are in agreements with the theoretical predictions. Subsequently, we perform numerical simulations to show that these results extend to stronger degrees of disorder. Our key finding is that, since two-photon correlations are only affected by the even-parity component of phase modulations, the number of independent adaptive optics elements required for optimizing the correlation can be effectively halved, offering a significant practical advantage in managing disorder in quantum systems., Comment: 11 pages, 9 figures

Traditional evaluation metrics like BLEU and ROUGE fall short when capturing the nuanced qualities of generated text, particularly when there is no single ground truth. In this paper, we explore the potential of Large Language Models (LLMs), specifically Google Gemini 1, to serve as automatic evaluators for non-standardized metrics in summarization and dialog-based tasks. We conduct experiments across multiple prompting strategies to examine how LLMs fare as quality evaluators when compared with human judgments on the SummEval and USR datasets, asking the model to generate both a score as well as a justification for the score. Furthermore, we explore the robustness of the LLM evaluator by using perturbed inputs. Our findings suggest that while LLMs show promise, their alignment with human evaluators is limited, they are not robust against perturbations and significant improvements are required for their standalone use as reliable evaluators for subjective metrics., Comment: Accepted at ICON 2024

Appearing in the modular transformation formula for the classical theta function $\theta_3(z)$ is the sum $S(h,k):=\sum_{j=1}^{k-1}(-1)^{j+1+[hj/k]}$, which is an analogue of the classical Dedekind sum $s(h,k).$ We establish several properties for $S(h,k)$ and $S(k) := \sum_{h=1}^{k-1}S(h,k).$ Several conjectures about the values of $S(k)$ are given.

In various video-language learning tasks, the challenge of achieving cross-modality alignment with multi-grained data persists. We propose a method to tackle this challenge from two crucial perspectives: data and modeling. Given the absence of a multi-grained video-text pretraining dataset, we introduce a Granularity EXpansion (GEX) method with Integration and Compression operations to expand the granularity of a single-grained dataset. To better model multi-grained data, we introduce an Iterative Approximation Module (IAM), which embeds multi-grained videos and texts into a unified, low-dimensional semantic space while preserving essential information for cross-modal alignment. Furthermore, GEXIA is highly scalable with no restrictions on the number of video-text granularities for alignment. We evaluate our work on three categories of video tasks across seven benchmark datasets, showcasing state-of-the-art or comparable performance. Remarkably, our model excels in tasks involving long-form video understanding, even though the pretraining dataset only contains short video clips.

We have conducted a GPU accelerated reprocessing of $\sim 87\%$ of the archival data from the High Time Resolution Universe South Low Latitude (HTRU-S LowLat) pulsar survey by implementing a pulsar search pipeline that was previously used to reprocess the Parkes Multibeam pulsar survey (PMPS). We coherently searched the full 72-min observations of the survey with an acceleration search range up to $|50|\, \rm m\,s^{-2}$, which is most sensitive to binary pulsars experiencing nearly constant acceleration during 72 minutes of their orbital period. Here we report the discovery of 71 pulsars, including 6 millisecond pulsars (MSPs) of which five are in binary systems, and seven pulsars with very high dispersion measures (DM $>800 \, \rm pc \, cm^{-3}$). These pulsar discoveries largely arose by folding candidates to a much lower spectral signal-to-noise ratio than previous surveys, and exploiting the coherence of folding over the incoherent summing of the Fourier components to discover new pulsars as well as candidate classification techniques. We show that these pulsars could be fainter and on average more distant as compared to both the previously reported 100 HTRU-S LowLat pulsars and background pulsar population in the survey region. We have assessed the effectiveness of our search method and the overall pulsar yield of the survey. We show that through this reprocessing we have achieved the expected survey goals including the predicted number of pulsars in the survey region and discuss the major causes as to why these pulsars were missed in previous processings of the survey., Comment: 18 Pages, 12 figures, 9 tables. Accepted for publication in MNRAS

Here, we present a new thermomechanical geodynamic, numerical implementation that incorporates Maxwell viscoelastic rheology accounting for temperature-dependent power-law dislocation creep and pressure-sensitive, non-associated Drucker-Prager brittle failure, as well as for volumetric stresses and strains during viscoplastic flow, a departure from the traditional incompressible assumptions. In solving for energy conservation, we incorporate the heat source term resulting from irreversible mechanical deformations, which embodies viscoelastic and viscoplastic work, and by considering the total stress tensor and total inelastic strain rate tensors, including dilatant plasticity effects for lithospheric-scale applications, instead of only the shear terms as is usually assumed for incompressible materials. This form of the work term thus allows to consider, volumetric deformation and to couple the energy equation to the constitutive description, and hence the stress balance, via the evolving temperature field. Code design enables us to switch individual features of this general rheology ``on or off'' and thus to benchmark this implementation with published numerical experiments of crustal-scale shortening experiments. We investigate whether ``brittle-plastic'' compressibility can promote or inhibit localization of deformation and thermal evolution during compression for crustal, and upper mantle rheology. For both crustal-scale and lithospheric-scale experiments, we establish that the feedback from volumetric dissipation, while contributing to temperature increase along with shear dissipation, can potentially slow down heat production per unit time, depending on the choice of boundary conditions. Our new implementation can be used to address buckling problems and collision tectonics.

The proliferation of several streaming services in recent years has now made it possible for a diverse audience across the world to view the same media content, such as movies or TV shows. While translation and dubbing services are being added to make content accessible to the local audience, the support for making content accessible to people with different abilities, such as the Deaf and Hard of Hearing (DHH) community, is still lagging. Our goal is to make media content more accessible to the DHH community by generating sign language videos with synthetic signers that are realistic and expressive. Using the same signer for a given media content that is viewed globally may have limited appeal. Hence, our approach combines parametric modeling and generative modeling to generate realistic-looking synthetic signers and customize their appearance based on user preferences. We first retarget human sign language poses to 3D sign language avatars by optimizing a parametric model. The high-fidelity poses from the rendered avatars are then used to condition the poses of synthetic signers generated using a diffusion-based generative model. The appearance of the synthetic signer is controlled by an image prompt supplied through a visual adapter. Our results show that the sign language videos generated using our approach have better temporal consistency and realism than signing videos generated by a diffusion model conditioned only on text prompts. We also support multimodal prompts to allow users to further customize the appearance of the signer to accommodate diversity (e.g. skin tone, gender). Our approach is also useful for signer anonymization., Comment: Published in Proceedings of ECCV, Workshop on Assistive Computer Vision and Robotics, 2024

Consider the problem of determining the optimal applied electric field to drive a molecule from an initial state to a desired target state. For even moderately sized molecules, solving this problem directly using the exact equations of motion -- the time-dependent Schr\"odinger equation (TDSE) -- is numerically intractable. We present a solution of this problem within time-dependent Hartree-Fock (TDHF) theory, a mean field approximation of the TDSE. Optimality is defined in terms of minimizing the total control effort while maximizing the overlap between desired and achieved target states. We frame this problem as an optimization problem constrained by the nonlinear TDHF equations; we solve it using trust region optimization with gradients computed via a custom-built adjoint state method. For three molecular systems, we show that with very small neural network parametrizations of the control, our method yields solutions that achieve desired targets within acceptable constraints and tolerances., Comment: 8 pages, 2 figures

Amodal depth estimation aims to predict the depth of occluded (invisible) parts of objects in a scene. This task addresses the question of whether models can effectively perceive the geometry of occluded regions based on visible cues. Prior methods primarily rely on synthetic datasets and focus on metric depth estimation, limiting their generalization to real-world settings due to domain shifts and scalability challenges. In this paper, we propose a novel formulation of amodal depth estimation in the wild, focusing on relative depth prediction to improve model generalization across diverse natural images. We introduce a new large-scale dataset, Amodal Depth In the Wild (ADIW), created using a scalable pipeline that leverages segmentation datasets and compositing techniques. Depth maps are generated using large pre-trained depth models, and a scale-and-shift alignment strategy is employed to refine and blend depth predictions, ensuring consistency in ground-truth annotations. To tackle the amodal depth task, we present two complementary frameworks: Amodal-DAV2, a deterministic model based on Depth Anything V2, and Amodal-DepthFM, a generative model that integrates conditional flow matching principles. Our proposed frameworks effectively leverage the capabilities of large pre-trained models with minimal modifications to achieve high-quality amodal depth predictions. Experiments validate our design choices, demonstrating the flexibility of our models in generating diverse, plausible depth structures for occluded regions. Our method achieves a 69.5% improvement in accuracy over the previous SoTA on the ADIW dataset.

We conduct the first-ever Pulsar Polarization Array (PPA) analysis to detect the ultralight Axion-Like Dark Matter (ALDM) using the polarization data of 22 millisecond pulsars from the third data release of Parkes Pulsar Timing Array. As one of the major dark matter candidates, the ultralight ALDM exhibits a pronounced wave nature on astronomical scales and offers a promising solution to small-scale structure issues within local galaxies. While the linearly polarized pulsar light travels through the ALDM galactic halo, its position angle (PA) can be subject to an oscillation induced by the ALDM Chern-Simons coupling with electromagnetic field. The PPA is thus especially suited for detecting the ultralight ALDM by correlating polarization data across the arrayed pulsars. To accomplish this task, we develop an advanced Bayesian analysis framework that allows us to construct pulsar PA residual time series, model noise contributions properly and search for pulsar cross-correlations. We find that for an ALDM density of $\rho_0=0.4\,\textrm{GeV}/\textrm{cm}^3$, the Parkes PPA offers the best global limits on the ALDM Chern-Simons coupling, namely $\lesssim 10^{-13.5}-10^{-12.2}~{\rm GeV}^{-1}$, for the mass range of $10^{-22} - 10^{-21}~{\rm eV}$. The crucial role of pulsar cross-correlation in recognizing the nature of the derived limits is also highlighted., Comment: 6+15 pages, 10 figures, 2 tables, submitted to the journal

Open-universe 3D layout generation arranges unlabeled 3D assets conditioned on language instruction. Large language models (LLMs) struggle with generating physically plausible 3D scenes and adherence to input instructions, particularly in cluttered scenes. We introduce LayoutVLM, a framework and scene layout representation that exploits the semantic knowledge of Vision-Language Models (VLMs) and supports differentiable optimization to ensure physical plausibility. LayoutVLM employs VLMs to generate two mutually reinforcing representations from visually marked images, and a self-consistent decoding process to improve VLMs spatial planning. Our experiments show that LayoutVLM addresses the limitations of existing LLM and constraint-based approaches, producing physically plausible 3D layouts better aligned with the semantic intent of input language instructions. We also demonstrate that fine-tuning VLMs with the proposed scene layout representation extracted from existing scene datasets can improve performance., Comment: project website: https://ai.stanford.edu/~sunfanyun/layoutvlm/

Let $P(z)$ be a polynomial of degree $n,$ then it is known that for $\alpha\in\mathbb{C}$ with $|\alpha|\leq \frac{n}{2},$ \begin{align*} \underset{|z|=1}{\max}|\left|zP^{\prime}(z)-\alpha P(z)\right|\leq \left|n-\alpha\right|\underset{|z|=1}{\max}|P(z)|. \end{align*} This inequality includes Bernstein's inequality, concerning the estimate for $|P^\prime(z)|$ over $|z|\leq 1,$ as a special case. In this paper, we extend this inequality to $L_p$ norm which among other things shows that the condition on $\alpha$ can be relaxed. We also prove similar inequalities for polynomials with restricted zeros., Comment: L^{p}$-inequalities, Bernstein's inequality, polynomials

In this review paper, we explore operator aspects in extremal properties of Bernstein-type polynomial inequalities. We shall also see that a linear operator which send polynomials to polynomials and have zero-preserving property naturally preserve Bernstein's inequality., Comment: Polynomials, Inequalities in complex domain, Bernstein's Inequality, Gauss-Lucas theorem

Multi-modal Large Language Models (MLLMs) are gaining significant attention for their ability to process multi-modal data, providing enhanced contextual understanding of complex problems. MLLMs have demonstrated exceptional capabilities in tasks such as Visual Question Answering (VQA); however, they often struggle with fundamental engineering problems, and there is a scarcity of specialized datasets for training on topics like digital electronics. To address this gap, we propose a benchmark dataset called ElectroVizQA specifically designed to evaluate MLLMs' performance on digital electronic circuit problems commonly found in undergraduate curricula. This dataset, the first of its kind tailored for the VQA task in digital electronics, comprises approximately 626 visual questions, offering a comprehensive overview of digital electronics topics. This paper rigorously assesses the extent to which MLLMs can understand and solve digital electronic circuit questions, providing insights into their capabilities and limitations within this specialized domain. By introducing this benchmark dataset, we aim to motivate further research and development in the application of MLLMs to engineering education, ultimately bridging the performance gap and enhancing the efficacy of these models in technical fields.