Search

Real-time Magnetic Resonance Imaging (rtMRI) is frequently used in speech production studies as it provides a complete view of the vocal tract during articulation. This study investigates the effectiveness of rtMRI in analyzing vocal tract movements by employing the SegNet and UNet models for Air-Tissue Boundary (ATB)segmentation tasks. We conducted pretraining of a few base models using increasing numbers of subjects and videos, to assess performance on two datasets. First, consisting of unseen subjects with unseen videos from the same data source, achieving 0.33% and 0.91% (Pixel-wise Classification Accuracy (PCA) and Dice Coefficient respectively) better than its matched condition. Second, comprising unseen videos from a new data source, where we obtained an accuracy of 99.63% and 98.09% (PCA and Dice Coefficient respectively) of its matched condition performance. Here, matched condition performance refers to the performance of a model trained only on the test subjects which was set as a benchmark for the other models. Our findings highlight the significance of fine-tuning and adapting models with limited data. Notably, we demonstrated that effective model adaptation can be achieved with as few as 15 rtMRI frames from any new dataset., Comment: Accepted to ICASSP 2025

This paper analyzes a random walk model for the level lines appearing in the entropic repulsion phenomena of three-dimensional discrete random interfaces above a hard wall; we are particularly motivated by the low-temperature (2+1)D solid-on-solid (SOS) model, where the emergence of these level lines has been rigorously established. The model we consider is a line ensemble of non-crossing random walk bridges above a wall with geometrically growing area tilts. Our main result, which in particular resolves a question of Caputo, Ioffe, and Wachtel (2019), is an edge 1:2:3 scaling limit for this ensemble as the domain size $N$ diverges, with a growing number of walks (including the number of level lines of the SOS model) and high boundary conditions (covering the maximum upper deviation of the SOS level lines). As a key input, we establish Tracy--Widom-type upper tail bounds for each of the relevant curves in the line ensemble. An ingredient which may be of independent interest is a ballot theorem for random walk bridges under a broader range of boundary values than available in the literature., Comment: 60 pages, 7 figures

Large reasoning models (LRMs) tackle complex reasoning problems by following long chain-of-thoughts (Long CoT) that incorporate reflection, backtracking, and self-validation. However, the training techniques and data requirements to elicit Long CoT remain poorly understood. In this work, we find that a Large Language model (LLM) can effectively learn Long CoT reasoning through data-efficient supervised fine-tuning (SFT) and parameter-efficient low-rank adaptation (LoRA). With just 17k long CoT training samples, the Qwen2.5-32B-Instruct model achieves significant improvements on a wide range of math and coding benchmarks, including 56.7% (+40.0%) on AIME 2024 and 57.0% (+8.1%) on LiveCodeBench, competitive to the proprietary o1-preview model's score of 44.6% and 59.1%. More importantly, we find that the structure of Long CoT is critical to the learning process, whereas the content of individual reasoning steps has minimal impact. Perturbations affecting content, such as training on incorrect samples or removing reasoning keywords, have little impact on performance. In contrast, structural modifications that disrupt logical consistency in the Long CoT, such as shuffling or deleting reasoning steps, significantly degrade accuracy. For example, a model trained on Long CoT samples with incorrect answers still achieves only 3.2% lower accuracy compared to training with fully correct samples. These insights deepen our understanding of how to elicit reasoning capabilities in LLMs and highlight key considerations for efficiently training the next generation of reasoning models. This is the academic paper of our previous released Sky-T1-32B-Preview model. Codes are available at https://github.com/NovaSky-AI/SkyThought.

Multiphase fluid dynamics, such as falling droplets and rising bubbles, are critical to many industrial applications. However, simulating these phenomena efficiently is challenging due to the complexity of instabilities, wave patterns, and bubble breakup. This paper investigates the potential of scientific machine learning (SciML) to model these dynamics using neural operators and foundation models. We apply sequence-to-sequence techniques on a comprehensive dataset generated from 11,000 simulations, comprising 1 million time snapshots, produced with a well-validated Lattice Boltzmann method (LBM) framework. The results demonstrate the ability of machine learning models to capture transient dynamics and intricate fluid interactions, paving the way for more accurate and computationally efficient SciML-based solvers for multiphase applications.

Recent advancements in human preference optimization, originally developed for Large Language Models (LLMs), have shown significant potential in improving text-to-image diffusion models. These methods aim to learn the distribution of preferred samples while distinguishing them from less preferred ones. However, existing preference datasets often exhibit overlap between these distributions, leading to a conflict distribution. Additionally, we identified that input prompts contain irrelevant information for less preferred images, limiting the denoising network's ability to accurately predict noise in preference optimization methods, known as the irrelevant prompt issue. To address these challenges, we propose Dual Caption Preference Optimization (DCPO), a novel approach that utilizes two distinct captions to mitigate irrelevant prompts. To tackle conflict distribution, we introduce the Pick-Double Caption dataset, a modified version of Pick-a-Pic v2 with separate captions for preferred and less preferred images. We further propose three different strategies for generating distinct captions: captioning, perturbation, and hybrid methods. Our experiments show that DCPO significantly improves image quality and relevance to prompts, outperforming Stable Diffusion (SD) 2.1, SFT_Chosen, Diffusion-DPO, and MaPO across multiple metrics, including Pickscore, HPSv2.1, GenEval, CLIPscore, and ImageReward, fine-tuned on SD 2.1 as the backbone.

As they become more capable, large language models (LLMs) have continued to rapidly increase in size. This has exacerbated the difficulty in running state of the art LLMs on small, edge devices. Standard techniques advocate solving this problem through lossy compression techniques such as quantization or pruning. However, such compression techniques are lossy, and have been shown to change model behavior in unpredictable manners. We propose Huff-LLM, an \emph{end-to-end, lossless} model compression method that lets users store LLM weights in compressed format \emph{everywhere} -- cloud, disk, main memory, and even in on-chip memory/buffers. This allows us to not only load larger models in main memory, but also reduces bandwidth required to load weights on chip, and makes more efficient use of on-chip weight buffers. In addition to the memory savings achieved via compression, we also show latency and energy efficiency improvements when performing inference with the compressed model.

Interplanetary Coronal Mass Ejections (ICMEs) are the primary sources of geomagnetic storms at Earth. Negative out-of-ecliptic component (Bz) of magnetic field in the ICME or its associated sheath region is necessary for it to be geo-effective. For this reason, magnetohydrodynamic simulations of CMEs containing data-constrained flux ropes are more suitable for forecasting their geo-effectiveness as compared to hydrodynamic models of the CME. ICMEs observed in situ by radially aligned spacecraft can provide an important setup to validate the physics-based heliospheric modeling of CMEs. In this work, we use the constant-turn flux rope (CTFR) model to study an ICME that was observed in situ by Solar Orbiter (SolO) and at Earth, when they were in a near-radial alignment. This was a stealth CME that erupted on 2020 April 14 and reached Earth on 2020 April 20 with a weak shock and a smoothly rotating magnetic field signature. We found that the CTFR model was able to reproduce the rotating magnetic field signature at both SolO and Earth with very good accuracy. The simulated ICME arrived 5 hours late at SolO and 5 hours ahead at Earth, when compared to the observed ICME. We compare the propagation of the CME front through the inner heliosphere using synthetic J-maps and those observed in the heliospheric imager data and discuss the role of incorrect ambient SW background on kinematics of the simulated CME. This study supports the choice of the CTFR model for reproducing the magnetic field of ICMEs.

Language model (LLM) post-training, from DPO to distillation, can refine behaviors and unlock new skills, but the open science supporting these post-training techniques is still in its infancy. One limiting factor has been the difficulty of conducting large-scale comparative analyses of synthetic data generating models and LLM judges. To close this gap, we introduce WILDCHAT-50M, the largest public chat dataset to date. We extend the existing WildChat dataset to include responses not only from GPT, but from over 50 different open-weight models, ranging in size from 0.5B to 104B parameters. We conduct an extensive comparative analysis and demonstrate the potential of this dataset by creating RE-WILD, our own public SFT mix, which outperforms the recent Tulu-3 SFT mixture from Allen AI with only 40% as many samples. Our dataset, samples and code are available at https://github.com/penfever/wildchat-50m.

Vision-language model (VLM) fine-tuning for application-specific visual grounding based on natural language instructions has become one of the most popular approaches for learning-enabled autonomous systems. However, such fine-tuning relies heavily on high-quality datasets to achieve successful performance in various downstream tasks. Additionally, VLMs often encounter limitations due to insufficient and imbalanced fine-tuning data. To address these issues, we propose a new generalizable framework to improve VLM fine-tuning by integrating it with a reinforcement learning (RL) agent. Our method utilizes the RL agent to manipulate objects within an indoor setting to create synthetic data for fine-tuning to address certain vulnerabilities of the VLM. Specifically, we use the performance of the VLM to provide feedback to the RL agent to generate informative data that efficiently fine-tune the VLM over the targeted task (e.g. spatial reasoning). The key contribution of this work is developing a framework where the RL agent serves as an informative data sampling tool and assists the VLM in order to enhance performance and address task-specific vulnerabilities. By targeting the data sampling process to address the weaknesses of the VLM, we can effectively train a more context-aware model. In addition, generating synthetic data allows us to have precise control over each scene and generate granular ground truth captions. Our results show that the proposed data generation approach improves the spatial reasoning performance of VLMs, which demonstrates the benefits of using RL-guided data generation in vision-language tasks., Comment: ICCPS 2025 accepted paper, 10 pages, 9 figures

Organometallic halide perovskites have garnered considerable attention in recent times due to their promising optoelectronic attributes, particularly within the realm of solar photovoltaics (PV). How perovskite films form is of utmost significance in shaping their structural and functional characteristics. In this context, the application of methylamine vapour during the precursor deposition and subsequent treatment during the film formation stages emerges as crucial for the development of high-quality perovskite films for solar cell applications. The utilization of methylamine vapour annealing is pivotal in improving the crystallinity, morphology, and overall integrity of perovskite films. This work investigates the characteristics of perovskite films based on methylamine lead iodide, focusing on aspects such as crystallographic structure and vibrational modes, which are directly linked to the performance of the devices. The maximum power conversion efficiencies (PCE) obtained are 19.5% and 18.6% using 1-step and 2-step processes are obtained. The effect of factors like trap states, film homogeneity, and interfaces on the device performance are explored through capacitance measurements, photoluminescence, and electroluminescence behaviour. The recombination behaviour of the perovskite films is correlated with the crystallographic properties. These findings provide valuable insights into the influence of different processing techniques, such as methylamine vapour treatment and vacuum annealing, on rejuvenating perovskite solar cells.

Autonomous driving demands safe motion planning, especially in critical "long-tail" scenarios. Recent end-to-end autonomous driving systems leverage large language models (LLMs) as planners to improve generalizability to rare events. However, using LLMs at test time introduces high computational costs. To address this, we propose DiMA, an end-to-end autonomous driving system that maintains the efficiency of an LLM-free (or vision-based) planner while leveraging the world knowledge of an LLM. DiMA distills the information from a multi-modal LLM to a vision-based end-to-end planner through a set of specially designed surrogate tasks. Under a joint training strategy, a scene encoder common to both networks produces structured representations that are semantically grounded as well as aligned to the final planning objective. Notably, the LLM is optional at inference, enabling robust planning without compromising on efficiency. Training with DiMA results in a 37% reduction in the L2 trajectory error and an 80% reduction in the collision rate of the vision-based planner, as well as a 44% trajectory error reduction in longtail scenarios. DiMA also achieves state-of-the-art performance on the nuScenes planning benchmark.

This article reports the development of a versatile high-speed reciprocating drive system (HRDS) with interchangeable probe heads to characterize the edge plasma region of ADITYA-U tokamak. This reciprocating probe drive system consisting of Langmuir and magnetic probe heads, is designed, fabricated, installed, and operated for studying the extent of fuel/impurity gas propagation and its influence on plasma dynamics in the far-edge region inside the last closed magnetic flux surface (LCFS). The HRDS is driven by a highly accurate, easy-to-control, dynamic, brushless, permanently excited synchronous servo motor operated by a PXI-commanded controller. The system is remotely operated and allows for precise control of the speed, acceleration, and distance traveled of the probe head on a shot-to-shot basis, facilitating seamless control of operations according to experimental requirements. Using this system, consisting of a linear array of Langmuir probes, measurements of plasma density, temperature, potential, and their fluctuations revealed that the fuel gas-puff impact these mean and fluctuating parameters up to three to four cm inside the LCFS. Attaching an array of magnetic probes to this system led to measurements of magnetic fluctuations inside the LCFS. The HRDS system is fully operational and serves as an important diagnostic tool for ADITYA-U tokamak.

Experiments on ADITYA-U tokamak show a marked enhancement in the sawtooth period by application of short gas puffs of fuel that cause a modification of the radial density profile. A consequent suppression of the trapped electron modes (TEMs) then leads to an increase in the core electron temperature. This slows down the heat propagation following a sawtooth crash, causing a delay in achieving the critical temperature gradient inside the q = 1 surface required for the next sawtooth crash to happen. The overall scenario has strong similarities with the behavior of sawtooth under electron cyclotron resonance heating (ECRH). Our findings suggest an alternate, simpler technique for sawtooth control that may be usefully employed in small/medium-sized tokamaks that do not have an ECRH or any other auxiliary heating facility.

Rapid yet accurate simulations of fluid dynamics around complex geometries is critical in a variety of engineering and scientific applications, including aerodynamics and biomedical flows. However, while scientific machine learning (SciML) has shown promise, most studies are constrained to simple geometries, leaving complex, real-world scenarios underexplored. This study addresses this gap by benchmarking diverse SciML models, including neural operators and vision transformer-based foundation models, for fluid flow prediction over intricate geometries. Using a high-fidelity dataset of steady-state flows across various geometries, we evaluate the impact of geometric representations -- Signed Distance Fields (SDF) and binary masks -- on model accuracy, scalability, and generalization. Central to this effort is the introduction of a novel, unified scoring framework that integrates metrics for global accuracy, boundary layer fidelity, and physical consistency to enable a robust, comparative evaluation of model performance. Our findings demonstrate that foundation models significantly outperform neural operators, particularly in data-limited scenarios, and that SDF representations yield superior results with sufficient training data. Despite these advancements, all models struggle with out-of-distribution generalization, highlighting a critical challenge for future SciML applications. By advancing both evaluation methodologies and modeling capabilities, this work paves the way for robust and scalable ML solutions for fluid dynamics across complex geometries.

The COVID-19 vaccine development, manufacturing, transportation, and administration proved an extreme logistics operation of global magnitude. Global vaccination levels, however, remain a key concern in preventing the emergence of new strains and minimizing the impact of the pandemic's disruption of daily life. In this paper, country-level vaccination rates are analyzed through a queuing framework to extract service rates that represent the practical capacity of a country to administer vaccines. These rates are further characterized through regression and interpretable machine learning methods with country-level demographic, governmental, and socio-economic variates. Model results show that participation in multi-governmental collaborations such as COVAX may improve the ability to vaccinate. Similarly, improved transportation and accessibility variates such as roads per area for low-income countries and rail lines per area for high-income countries can improve rates. It was also found that for low-income countries specifically, improvements in basic and health infrastructure (as measured through spending on healthcare, number of doctors and hospital beds per 100k, population percent with access to electricity, life expectancy, and vehicles per 1000 people) resulted in higher vaccination rates. Of the high-income countries, those with larger 65-plus populations struggled to vaccinate at high rates, indicating potential accessibility issues for the elderly. This study finds that improving basic and health infrastructure, focusing on accessibility in the last mile, particularly for the elderly, and fostering global partnerships can improve logistical operations of such a scale. Such structural impediments and inequities in global health care must be addressed in preparation for future global public health crises., Comment: Under consideration for more thorough analysis

We present STITCH, a novel approach for neural implicit surface reconstruction of a sparse and irregularly spaced point cloud while enforcing topological constraints (such as having a single connected component). We develop a new differentiable framework based on persistent homology to formulate topological loss terms that enforce the prior of a single 2-manifold object. Our method demonstrates excellent performance in preserving the topology of complex 3D geometries, evident through both visual and empirical comparisons. We supplement this with a theoretical analysis, and provably show that optimizing the loss with stochastic (sub)gradient descent leads to convergence and enables reconstructing shapes with a single connected component. Our approach showcases the integration of differentiable topological data analysis tools for implicit surface reconstruction., Comment: 19 pages, 12 figures, 29 tables

We consider the stochastic six-vertex (S6V) model and asymmetric simple exclusion process (ASEP) under general initial conditions which are bounded below lines of arbitrary slope at $\pm\infty$. We show under Kardar-Parisi-Zhang (KPZ) scaling of time, space, and fluctuations that the height functions of these models converge to the KPZ fixed point. Previously, our results were known in the case of ASEP (for a particular direction in the rarefaction fan) via a comparison approach arXiv:2008.06584., Comment: 37 pages, 8 figures

Radio remains a pervasive medium for mass information dissemination, with AM/FM stations reaching more Americans than either smartphone-based social networking or live television. Increasingly, radio broadcasts are also streamed online and accessed over the Internet. We present WavePulse, a framework that records, documents, and analyzes radio content in real-time. While our framework is generally applicable, we showcase the efficacy of WavePulse in a collaborative project with a team of political scientists focusing on the 2024 Presidential Elections. We use WavePulse to monitor livestreams of 396 news radio stations over a period of three months, processing close to 500,000 hours of audio streams. These streams were converted into time-stamped, diarized transcripts and analyzed to track answer key political science questions at both the national and state levels. Our analysis revealed how local issues interacted with national trends, providing insights into information flow. Our results demonstrate WavePulse's efficacy in capturing and analyzing content from radio livestreams sourced from the Web. Code and dataset can be accessed at \url{https://wave-pulse.io}., Comment: To appear at The Web Conference (WWW) 2025. 20 Pages, 24 figures. Access code and dataset at https://wave-pulse.io

We consider the stationary measure of the asymmetric simple exclusion process (ASEP) on a finite interval in $\mathbb{Z}$ with open boundaries. Fixing all the jump rates and letting the system size approach infinity, the height profile of such a sequence of stationary measures satisfies a large deviation principle (LDP), whose rate function was predicted in the physics work arXiv:cond-mat/0205353. In this paper, we provide the first rigorous proof of the large deviation principle in the "fan region" part of the phase diagram. Our proof relies on two key ingredients: a two-layer expression of the stationary measure of open ASEP, arising from the Enaud-Derrida representation arXiv:cond-mat/0307023 of the matrix product ansatz, and the large deviation principle of the open totally asymmetric simple exclusion process (TASEP) recently established in arXiv:2403.03275., Comment: 19 pages

Supplementing the Heisenberg model with a Hubbard-commuting kinetic of electrons adds to its spectrum without interference. One consequence is the precise incorporation of canonical linear spin wave theory within the time-dependent Hartree-Fock framework, as pure localization emerges from itinerant dynamics. This embedding method generalizes to all spin-1/2 models and is expected to extend to multi-orbital systems. Away from half-filling, differential tuning of doublon and holon motion imparts asymmetry to ordering and fluctuations. This suggests that, in effective electronic theories, kinetic interaction couplings are as significant as underlying band parameters when modeling asymmetric phenomena near the Mott insulator., Comment: 7 pages, 4 figures

As the quality of image generators continues to improve, deepfakes become a topic of considerable societal debate. Image watermarking allows responsible model owners to detect and label their AI-generated content, which can mitigate the harm. Yet, current state-of-the-art methods in image watermarking remain vulnerable to forgery and removal attacks. This vulnerability occurs in part because watermarks distort the distribution of generated images, unintentionally revealing information about the watermarking techniques. In this work, we first demonstrate a distortion-free watermarking method for images, based on a diffusion model's initial noise. However, detecting the watermark requires comparing the initial noise reconstructed for an image to all previously used initial noises. To mitigate these issues, we propose a two-stage watermarking framework for efficient detection. During generation, we augment the initial noise with generated Fourier patterns to embed information about the group of initial noises we used. For detection, we (i) retrieve the relevant group of noises, and (ii) search within the given group for an initial noise that might match our image. This watermarking approach achieves state-of-the-art robustness to forgery and removal against a large battery of attacks.

In India, the majority of farmers are classified as small or marginal, making their livelihoods particularly vulnerable to economic losses due to market saturation and climate risks. Effective crop planning can significantly impact their expected income, yet existing decision support systems (DSS) often provide generic recommendations that fail to account for real-time market dynamics and the interactions among multiple farmers. In this paper, we evaluate the viability of three multi-agent reinforcement learning (MARL) approaches for optimizing total farmer income and promoting fairness in crop planning: Independent Q-Learning (IQL), where each farmer acts independently without coordination, Agent-by-Agent (ABA), which sequentially optimizes each farmer's policy in relation to the others, and the Multi-agent Rollout Policy, which jointly optimizes all farmers' actions for global reward maximization. Our results demonstrate that while IQL offers computational efficiency with linear runtime, it struggles with coordination among agents, leading to lower total rewards and an unequal distribution of income. Conversely, the Multi-agent Rollout policy achieves the highest total rewards and promotes equitable income distribution among farmers but requires significantly more computational resources, making it less practical for large numbers of agents. ABA strikes a balance between runtime efficiency and reward optimization, offering reasonable total rewards with acceptable fairness and scalability. These findings highlight the importance of selecting appropriate MARL approaches in DSS to provide personalized and equitable crop planning recommendations, advancing the development of more adaptive and farmer-centric agricultural decision-making systems.

Private inference (PI) serves an important role in guaranteeing the privacy of user data when interfacing with proprietary machine learning models such as LLMs. However, PI remains practically intractable due to the massive latency costs associated with nonlinear functions present in LLMs. Existing works have focused on improving latency of specific LLM nonlinearities (such as the Softmax, or the GeLU) via approximations. However, new types of nonlinearities are regularly introduced with new LLM architectures, and this has led to a constant game of catch-up where PI researchers attempt to optimize the newest nonlinear function. We introduce TruncFormer, a framework for taking any LLM and transforming it into a plaintext emulation of PI. Our framework leverages the fact that nonlinearities in LLMs are differentiable and can be accurately approximated with a sequence of additions, multiplications, and truncations. Further, we decouple the add/multiply and truncation operations, and statically determine where truncations should be inserted based on a given field size and input representation size. This leads to latency improvements over existing cryptographic protocols that enforce truncation after every multiplication operation. We open source our code for community use.

In autonomous driving, environment perception has significantly advanced with the utilization of deep learning techniques for diverse sensors such as cameras, depth sensors, or infrared sensors. The diversity in the sensor stack increases the safety and contributes to robustness against adverse weather and lighting conditions. However, the variance in data acquired from different sensors poses challenges. In the context of continual learning (CL), incremental learning is especially challenging for considerably large domain shifts, e.g. different sensor modalities. This amplifies the problem of catastrophic forgetting. To address this issue, we formulate the concept of modality-incremental learning and examine its necessity, by contrasting it with existing incremental learning paradigms. We propose the use of a modified Relevance Mapping Network (RMN) to incrementally learn new modalities while preserving performance on previously learned modalities, in which relevance maps are disjoint. Experimental results demonstrate that the prevention of shared connections in this approach helps alleviate the problem of forgetting within the constraints of a strict continual learning framework., Comment: Accepted at WACV 2025

Video understanding has witnessed significant progress with recent video foundation models demonstrating strong performance owing to self-supervised pre-training objectives; Masked Autoencoders (MAE) being the design of choice. Nevertheless, the majority of prior works that leverage MAE pre-training have focused on relatively short video representations (16 / 32 frames in length) largely due to hardware memory and compute limitations that scale poorly with video length due to the dense memory-intensive self-attention decoding. One natural strategy to address these challenges is to subsample tokens to reconstruct during decoding (or decoder masking). In this work, we propose an effective strategy for prioritizing tokens which allows training on longer video sequences (128 frames) and gets better performance than, more typical, random and uniform masking strategies. The core of our approach is an adaptive decoder masking strategy that prioritizes the most important tokens and uses quantized tokens as reconstruction objectives. Our adaptive strategy leverages a powerful MAGVIT-based tokenizer that jointly learns the tokens and their priority. We validate our design choices through exhaustive ablations and observe improved performance of the resulting long-video (128 frames) encoders over short-video (32 frames) counterparts. With our long-video masked autoencoder (LVMAE) strategy, we surpass state-of-the-art on Diving48 by 3.9 points and EPIC-Kitchens-100 verb classification by 2.5 points while relying on a simple core architecture and video-only pre-training (unlike some of the prior works that require millions of labeled video-text pairs or specialized encoders)., Comment: 10.5 pages of main paper, 25 pages total, 4 figures and 10 tables. To appear in NeurIPS'24

Our research aims to probe the anisotropic matter field around black holes using black hole perturbation theory. Black holes in the universe are usually surrounded by matter or fields, and it is important to study the perturbation and the characteristic modes of a black hole that coexists with such a matter field. In this study, we focus on a family of black hole solutions to Einstein's equations that extend the Reissner-Nordstr\"{o}m spacetime to include an anisotropic matter field. In addition to mass and charge, this type of black hole possesses additional hair due to the negative radial pressure of the anisotropic matter. We investigate the perturbations of the massless scalar and electromagnetic fields and calculate the quasinormal modes (QNMs). We also study the critical orbits around the black hole and their properties to investigate the connection between the eikonal QNMs, black hole shadow radius, and Lyapunov exponent. Additionally, we analyze the grey-body factors and scattering coefficients using the perturbation results. Our findings indicate that the presence of anisotropic matter fields leads to a splitting in the QNM frequencies compared to the Schwarzschild case. This splitting feature is also reflected in the shadow radius, Lyapunov exponent, and grey-body factors., Comment: 35 pages, 10 figures

The differential representation of AdS correlators offers a framework to express exchange Witten diagrams as functions of non-local differential operators applied to contact Witten diagrams. In this paper, we develop the differential representation for scalar Carrollian correlators. We first construct this representation using the recently formulated Carrollian limit of AdS Witten diagrams. We then provide an alternate intrinsic analysis that leverages the properties of the Carrollian bulk-to-boundary propagator. Using the differential representation, we also obtain differential Bern-Carrasco-Johansson (BCJ) relations for Carrollian correlators., Comment: 27 pages, 3 figures, 2 tables

We compute the supersymmetric index of half BPS black holes in N=2 supergravity with higher curvature corrections and show that the result agrees with the degeneracy of supersymmetric extremal black holes carrying the same charges. Both sides of the computation are done gravitationally., Comment: 27 pages

Satellites are widely used to estimate and monitor ground cover, providing critical information to address the challenges posed by climate change. High-resolution satellite images help to identify smaller features on the ground and classification of ground cover types. Small satellites have become very popular recently due to their cost-effectiveness. However, smaller satellites have weaker spatial resolution, and preprocessing using recent generative models made it possible to enhance the resolution of these satellite images. The objective of this paper is to propose computationally efficient guided or image-conditioned denoising diffusion models (DDMs) to perform super-resolution on low-quality images. Denoising based on stochastic ordinary differential equations (ODEs) typically takes hundreds of iterations and it can be reduced using deterministic ODEs. I propose Consistency Models (CM) that utilize deterministic ODEs for efficient denoising and perform super resolution on satellite images. The DOTA v2.0 image dataset that is used to develop object detectors needed for urban planning and ground cover estimation, is used in this project. The Stable Diffusion model is used as the base model, and the DDM in Stable Diffusion is converted into a Consistency Model (CM) using Teacher-Student Distillation to apply deterministic denoising. Stable diffusion with modified CM has successfully improved the resolution of satellite images by a factor of 16, and the computational time was reduced by a factor of 20 compared to stochastic denoising methods. The FID score of low-resolution images improved from 10.0 to 1.9 after increasing the image resolution using my algorithm for consistency models.

Erd\H{o}s and Gy\'arf\'as conjectured in 1994 that every graph with minimum degree at least 3 has a cycle of length a power of 2. In 2022, Gao and Shan (Graphs and Combinatorics) proved that the conjecture is true for $P_8$-free graphs, i.e., graphs without any induced copies of a path on 8 vertices. In 2024, Hu and Shen (Discrete Mathematics) improved this result by proving that the conjecture is true for $P_{10}$ -free graphs. With the aid of a computer search, we improve this further by proving that the conjecture is true for $P_{13}$ -free graphs., Comment: 6 pages

Effective data-driven biomedical discovery requires data curation: a time-consuming process of finding, organizing, distilling, integrating, interpreting, annotating, and validating diverse information into a structured form suitable for databases and knowledge bases. Accurate and efficient curation of these digital assets is critical to ensuring that they are FAIR, trustworthy, and sustainable. Unfortunately, expert curators face significant time and resource constraints. The rapid pace of new information being published daily is exceeding their capacity for curation. Generative AI, exemplified by instruction-tuned large language models (LLMs), has opened up new possibilities for assisting human-driven curation. The design philosophy of agents combines the emerging abilities of generative AI with more precise methods. A curator's tasks can be aided by agents for performing reasoning, searching ontologies, and integrating knowledge across external sources, all efforts otherwise requiring extensive manual effort. Our LLM-driven annotation tool, CurateGPT, melds the power of generative AI together with trusted knowledge bases and literature sources. CurateGPT streamlines the curation process, enhancing collaboration and efficiency in common workflows. Compared to direct interaction with an LLM, CurateGPT's agents enable access to information beyond that in the LLM's training data and they provide direct links to the data supporting each claim. This helps curators, researchers, and engineers scale up curation efforts to keep pace with the ever-increasing volume of scientific data.

Offline paired preference optimization algorithms have become a popular approach for fine-tuning on preference data, outperforming traditional supervised fine-tuning in various tasks. However, traditional implementations often involve redundant computations, especially for tasks with long shared prompts. We introduce prefix sharing for preference tuning, a novel technique that processes chosen and rejected responses as one sequence with a shared prefix. To prevent cross-response contamination, we use a custom block-sparse attention mask. Our method achieves $1.1$-$1.5\times$ improvement in training throughput on popular DPO datasets, without any effect on convergence. When combined with sequence packing, we observe consistent $1.3$-$1.6\times$ speedups, benefiting even datasets with smaller sequence lengths. While we focus on Direct Preference Optimization (DPO), our approach is applicable to other paired preference tuning methods. By enhancing computational efficiency, our work contributes to making preference-based fine-tuning more accessible for a wider range of applications and model sizes. We open-source our code at https://github.com/frankxwang/dpo-prefix-sharing., Comment: To appear in NeurIPS 2024 in the Fine-Tuning in Machine Learning Workshop

With the increasing availability of open-source robotic data, imitation learning has emerged as a viable approach for both robot manipulation and locomotion. Currently, large generalized policies are trained to predict controls or trajectories using diffusion models, which have the desirable property of learning multimodal action distributions. However, generalizability comes with a cost - namely, larger model size and slower inference. Further, there is a known trade-off between performance and action horizon for Diffusion Policy (i.e., diffusing trajectories): fewer diffusion queries accumulate greater trajectory tracking errors. Thus, it is common practice to run these models at high inference frequency, subject to robot computational constraints. To address these limitations, we propose Latent Weight Diffusion (LWD), a method that uses diffusion to learn a distribution over policies for robotic tasks, rather than over trajectories. Our approach encodes demonstration trajectories into a latent space and then decodes them into policies using a hypernetwork. We employ a diffusion denoising model within this latent space to learn its distribution. We demonstrate that LWD can reconstruct the behaviors of the original policies that generated the trajectory dataset. LWD offers the benefits of considerably smaller policy networks during inference and requires fewer diffusion model queries. When tested on the Metaworld MT10 benchmark, LWD achieves a higher success rate compared to a vanilla multi-task policy, while using models up to ~18x smaller during inference. Additionally, since LWD generates closed-loop policies, we show that it outperforms Diffusion Policy in long action horizon settings, with reduced diffusion queries during rollout.

Gaussian processes are notorious for scaling cubically with the size of the training set, preventing application to very large regression problems. Computation-aware Gaussian processes (CAGPs) tackle this scaling issue by exploiting probabilistic linear solvers to reduce complexity, widening the posterior with additional computational uncertainty due to reduced computation. However, the most commonly used CAGP framework results in (sometimes dramatically) conservative uncertainty quantification, making the posterior unrealistic in practice. In this work, we prove that if the utilised probabilistic linear solver is calibrated, in a rigorous statistical sense, then so too is the induced CAGP. We thus propose a new CAGP framework, CAGP-GS, based on using Gauss-Seidel iterations for the underlying probabilistic linear solver. CAGP-GS performs favourably compared to existing approaches when the test set is low-dimensional and few iterations are performed. We test the calibratedness on a synthetic problem, and compare the performance to existing approaches on a large-scale global temperature regression problem.

Data curation is the problem of how to collect and organize samples into a dataset that supports efficient learning. Despite the centrality of the task, little work has been devoted towards a large-scale, systematic comparison of various curation methods. In this work, we take steps towards a formal evaluation of data curation strategies and introduce SELECT, the first large-scale benchmark of curation strategies for image classification. In order to generate baseline methods for the SELECT benchmark, we create a new dataset, ImageNet++, which constitutes the largest superset of ImageNet-1K to date. Our dataset extends ImageNet with 5 new training-data shifts, each approximately the size of ImageNet-1K itself, and each assembled using a distinct curation strategy. We evaluate our data curation baselines in two ways: (i) using each training-data shift to train identical image classification models from scratch (ii) using the data itself to fit a pretrained self-supervised representation. Our findings show interesting trends, particularly pertaining to recent methods for data curation such as synthetic data generation and lookup based on CLIP embeddings. We show that although these strategies are highly competitive for certain tasks, the curation strategy used to assemble the original ImageNet-1K dataset remains the gold standard. We anticipate that our benchmark can illuminate the path for new methods to further reduce the gap. We release our checkpoints, code, documentation, and a link to our dataset at https://github.com/jimmyxu123/SELECT., Comment: NeurIPS 2024, Datasets and Benchmarks Track

Deep learning-based methods have been widely researched in the areas of language and vision, demonstrating their capacity to understand long sequences of data and their usefulness in numerous helio-physics applications. Foundation models (FMs), which are pre-trained on a large-scale datasets, form the basis for a variety of downstream tasks. These models, especially those based on transformers in vision and language, show exceptional potential for adapting to a wide range of downstream applications. In this paper, we provide our perspective on the criteria for designing an FM for heliophysics and associated challenges and applications using the Solar Dynamics Observatory (SDO) dataset. We believe that this is the first study to design an FM in the domain of heliophysics., Comment: 31 Pages, 12 figures

Simulating fluid flow around arbitrary shapes is key to solving various engineering problems. However, simulating flow physics across complex geometries remains numerically challenging and computationally resource-intensive, particularly when using conventional PDE solvers. Machine learning methods offer attractive opportunities to create fast and adaptable PDE solvers. However, benchmark datasets to measure the performance of such methods are scarce, especially for flow physics across complex geometries. We introduce FlowBench, a dataset for neural simulators with over 10K samples, which is currently larger than any publicly available flow physics dataset. FlowBench contains flow simulation data across complex geometries (\textit{parametric vs. non-parametric}), spanning a range of flow conditions (\textit{Reynolds number and Grashoff number}), capturing a diverse array of flow phenomena (\textit{steady vs. transient; forced vs. free convection}), and for both 2D and 3D. FlowBench contains over 10K data samples, with each sample the outcome of a fully resolved, direct numerical simulation using a well-validated simulator framework designed for modeling transport phenomena in complex geometries. For each sample, we include velocity, pressure, and temperature field data at 3 different resolutions and several summary statistics features of engineering relevance (such as coefficients of lift and drag, and Nusselt numbers). %Additionally, we include masks and signed distance fields for each shape. We envision that FlowBench will enable evaluating the interplay between complex geometry, coupled flow phenomena, and data sufficiency on the performance of current, and future, neural PDE solvers. We enumerate several evaluation metrics to help rank order the performance of neural PDE solvers. We benchmark the performance of several baseline methods including FNO, CNO, WNO, and DeepONet.

We propose the use of latent space generative world models to address the covariate shift problem in autonomous driving. A world model is a neural network capable of predicting an agent's next state given past states and actions. By leveraging a world model during training, the driving policy effectively mitigates covariate shift without requiring an excessive amount of training data. During end-to-end training, our policy learns how to recover from errors by aligning with states observed in human demonstrations, so that at runtime it can recover from perturbations outside the training distribution. Additionally, we introduce a novel transformer-based perception encoder that employs multi-view cross-attention and a learned scene query. We present qualitative and quantitative results, demonstrating significant improvements upon prior state of the art in closed-loop testing in the CARLA simulator, as well as showing the ability to handle perturbations in both CARLA and NVIDIA's DRIVE Sim., Comment: 7 pages, 6 figures, for ICRA 2025 conference, for associated video file, see https://youtu.be/fO7RZ57gVxk

Ontologies and knowledge graphs (KGs) are general-purpose computable representations of some domain, such as human anatomy, and are frequently a crucial part of modern information systems. Most of these structures change over time, incorporating new knowledge or information that was previously missing. Managing these changes is a challenge, both in terms of communicating changes to users, and providing mechanisms to make it easier for multiple stakeholders to contribute. To fill that need, we have created KGCL, the Knowledge Graph Change Language, a standard data model for describing changes to KGs and ontologies at a high level, and an accompanying human-readable controlled natural language. This language serves two purposes: a curator can use it to request desired changes, and it can also be used to describe changes that have already happened, corresponding to the concepts of "apply patch" and "diff" commonly used for managing changes in text documents and computer programs. Another key feature of KGCL is that descriptions are at a high enough level to be useful and understood by a variety of stakeholders--for example, ontology edits can be specified by commands like "add synonym 'arm' to 'forelimb'" or "move 'Parkinson disease' under 'neurodegenerative disease'". We have also built a suite of tools for managing ontology changes. These include an automated agent that integrates with and monitors GitHub ontology repositories and applies any requested changes, and a new component in the BioPortal ontology resource that allows users to make change requests directly from within the BioPortal user interface. Overall, the KGCL data model, its controlled natural language, and associated tooling allow for easier management and processing of changes associated with the development of ontologies and KGs.

The magnetic field strength and its topology play an important role in understanding the formation, evolution, and dynamics of the solar corona. Also, it plays a significant role in addressing long-standing mysteries such as coronal heating problem, origin and propagation of coronal mass ejections, drivers of space weather, origin and acceleration of solar wind, and so on. Despite having photospheric magnetograms for decades, we do not have reliable observations of coronal magnetic field strengths today. To measure the coronal magnetic field precisely, the spectropolarimetry channel of the Visible Emission Line Coronagraph (VELC) on board the Aditya-L1 mission is designed. Using the observations of coronal emission line Fe XIII [10747{\AA~}], it is possible to generate full Stokes maps (I, Q, U, and V) that help in estimating the Line-of-Sight (LOS) magnetic field strength and to derive the magnetic field topology maps of solar corona in the Field of View (FOV) (1.05 -- 1.5~R$_{\odot}$). In this article, we summarize the instrumental details of the spectropolarimetry channel and detailed calibration procedures adopted to derive the modulation and demodulation matrices. Furthermore, we have applied the derived demodulation matrices to the observed data in the laboratory and studied their performance., Comment: 12 pages, 5 Figures, Published in Journal of Experimental Astronomy

Recent developments in vision language models (VLM) have shown great potential for diverse applications related to image understanding. In this study, we have explored state-of-the-art VLM models for vision-based transportation engineering tasks such as image classification and object detection. The image classification task involves congestion detection and crack identification, whereas, for object detection, helmet violations were identified. We have applied open-source models such as CLIP, BLIP, OWL-ViT, Llava-Next, and closed-source GPT-4o to evaluate the performance of these state-of-the-art VLM models to harness the capabilities of language understanding for vision-based transportation tasks. These tasks were performed by applying zero-shot prompting to the VLM models, as zero-shot prompting involves performing tasks without any training on those tasks. It eliminates the need for annotated datasets or fine-tuning for specific tasks. Though these models gave comparative results with benchmark Convolutional Neural Networks (CNN) models in the image classification tasks, for object localization tasks, it still needs improvement. Therefore, this study provides a comprehensive evaluation of the state-of-the-art VLM models highlighting the advantages and limitations of the models, which can be taken as the baseline for future improvement and wide-scale implementation.

We demonstrate that the spinning LQG black hole can act as a cosmic particle accelerator. The LQG solution is singularity-free and can possess spin greater than that of a Kerr black hole. The additional black hole hair, arising from quantum effects, significantly influences the particle dynamics around the black hole. Under suitable physical conditions, the center-of-mass energy can grow arbitrarily high during the collision of two generic particles in the spacetime of an extremal black hole. In the non-extremal case, there exists a finite upper bound on the center-of-mass energy, the maximum value of which depends on the LQG parameter. These results are particularly interesting from an astrophysical perspective, especially in the context of probing Planck-scale physics., Comment: 19 pages, 6 figures