Authors: Zeqian Chen
Abstract: The introduction of the transformer architecture in 2017 (cf.\cite{VSP2017}) marked the most striking advancement in natural language processing. The transformer is a model architecture relying entirely on an attention mechanism to draw global dependencies between input and output. However, we believe there is a gap in our theoretical understanding of what the transformer is, and why it works physically. In this paper, from a physical perspective on modern chips, we construct physical models in the Fock space over the Hilbert space of tokens realizing large language models based on a transformer architecture as open quantum systems. Our physical models underlie the transformer architecture for large language models.
Authors: Charvi Rastogi, Tian Huey Teh, Pushkar Mishra, Roma Patel, Ding Wang, Mark D\'iaz, Alicia Parrish, Aida Mostafazadeh Davani, Zoe Ashwood, Michela Paganini, Vinodkumar Prabhakaran, Verena Rieser, Lora Aroyo
Abstract: Current text-to-image (T2I) models often fail to account for diverse human experiences, leading to misaligned systems. We advocate for pluralistic alignment, where an AI understands and is steerable towards diverse, and often conflicting, human values. Our work provides three core contributions to achieve this in T2I models. First, we introduce a novel dataset for Diverse Intersectional Visual Evaluation (DIVE) -- the first multimodal dataset for pluralistic alignment. It enable deep alignment to diverse safety perspectives through a large pool of demographically intersectional human raters who provided extensive feedback across 1000 prompts, with high replication, capturing nuanced safety perceptions. Second, we empirically confirm demographics as a crucial proxy for diverse viewpoints in this domain, revealing significant, context-dependent differences in harm perception that diverge from conventional evaluations. Finally, we discuss implications for building aligned T2I models, including efficient data collection strategies, LLM judgment capabilities, and model steerability towards diverse perspectives. This research offers foundational tools for more equitable and aligned T2I systems. Content Warning: The paper includes sensitive content that may be harmful.
Authors: Jakub Strawa, Jarek Duda
Abstract: Data normalization is crucial in machine learning, usually performed by subtracting the mean and dividing by standard deviation, or by rescaling to a fixed range. In copula theory, popular in finance, there is used normalization to approximately quantiles by transforming x to CDF(x) with estimated CDF (cumulative distribution function) to nearly uniform distribution in [0,1], allowing for simpler representations which are less likely to overfit. It seems nearly unknown in machine learning, therefore, we would like to present some its advantages on example of recently popular Kolmogorov-Arnold Networks (KANs), improving predictions from Legendre-KAN by just switching rescaling to CDF normalization. Additionally, in HCR interpretation, weights of such neurons are mixed moments providing local joint distribution models, allow to propagate also probability distributions, and change propagation direction.
Authors: Mert Sehri, Zehui Hua, Francisco de Assis Boldt, Patrick Dumond
Abstract: Deep learning has revolutionized many industries by enabling models to automatically learn complex patterns from raw data, reducing dependence on manual feature engineering. However, deep learning algorithms are sensitive to input data, and performance often deteriorates under nonstationary conditions and across dissimilar domains, especially when using time-domain data. Conventional single-channel or parallel multi-source data loading strategies either limit generalization or increase computational costs. This study introduces selective embedding, a novel data loading strategy, which alternates short segments of data from multiple sources within a single input channel. Drawing inspiration from cognitive psychology, selective embedding mimics human-like information processing to reduce model overfitting, enhance generalization, and improve computational efficiency. Validation is conducted using six time-domain datasets, demonstrating that the proposed method consistently achieves high classification accuracy across various deep learning architectures while significantly reducing training times. The approach proves particularly effective for complex systems with multiple data sources, offering a scalable and resource-efficient solution for real-world applications in healthcare, heavy machinery, marine, railway, and agriculture, where robustness and adaptability are critical.
Authors: Aleksey Lapin, Igor Hromov, Stanislav Chumakov, Mile Mitrovic, Dmitry Simakov, Nikolay O. Nikitin, Andrey V. Savchenko
Abstract: AutoML has advanced in handling complex tasks using the integration of LLMs, yet its efficiency remains limited by dependence on specific underlying tools. In this paper, we introduce LightAutoDS-Tab, a multi-AutoML agentic system for tasks with tabular data, which combines an LLM-based code generation with several AutoML tools. Our approach improves the flexibility and robustness of pipeline design, outperforming state-of-the-art open-source solutions on several data science tasks from Kaggle. The code of LightAutoDS-Tab is available in the open repository https://github.com/sb-ai-lab/LADS
Authors: Alexander Strunk, Roland Assam
Abstract: This paper introduces Gauge Flow Models, a novel class of Generative Flow Models. These models incorporate a learnable Gauge Field within the Flow Ordinary Differential Equation (ODE). A comprehensive mathematical framework for these models, detailing their construction and properties, is provided. Experiments using Flow Matching on Gaussian Mixture Models demonstrate that Gauge Flow Models yields significantly better performance than traditional Flow Models of comparable or even larger size. Additionally, unpublished research indicates a potential for enhanced performance across a broader range of generative tasks.
Authors: Jiaxiang Yi, Bernardo P. Ferreira, Miguel A. Bessa
Abstract: Data-driven learning is generalized to consider history-dependent multi-fidelity data, while quantifying epistemic uncertainty and disentangling it from data noise (aleatoric uncertainty). This generalization is hierarchical and adapts to different learning scenarios: from training the simplest single-fidelity deterministic neural networks up to the proposed multi-fidelity variance estimation Bayesian recurrent neural networks. The versatility and generality of the proposed methodology are demonstrated by applying it to different data-driven constitutive modeling scenarios that include multiple fidelities with and without aleatoric uncertainty (noise). The method accurately predicts the response and quantifies model error while also discovering the noise distribution (when present). This opens opportunities for future real-world applications in diverse scientific and engineering domains; especially, the most challenging cases involving design and analysis under uncertainty.
Authors: Armel Soubeiga (LIMOS), Thomas Guyet (AISTROSIGHT), Violaine Antoine (LIMOS)
Abstract: Clustering based on belief functions has been gaining increasing attention in the machine learning community due to its ability to effectively represent uncertainty and/or imprecision. However, none of the existing algorithms can be applied to complex data, such as mixed data (numerical and categorical) or non-tabular data like time series. Indeed, these types of data are, in general, not represented in a Euclidean space and the aforementioned algorithms make use of the properties of such spaces, in particular for the construction of barycenters. In this paper, we reformulate the Evidential C-Means (ECM) problem for clustering complex data. We propose a new algorithm, Soft-ECM, which consistently positions the centroids of imprecise clusters requiring only a semi-metric. Our experiments show that Soft-ECM present results comparable to conventional fuzzy clustering approaches on numerical data, and we demonstrate its ability to handle mixed data and its benefits when combining fuzzy clustering with semi-metrics such as DTW for time series data.
Authors: Edward Henderson, Dewi Gould, Richard Everson, George De Ath, Nick Pepper
Abstract: Real-time assessment of near-term Air Traffic Controller (ATCO) task demand is a critical challenge in an increasingly crowded airspace, as existing complexity metrics often fail to capture nuanced operational drivers beyond simple aircraft counts. This work introduces an interpretable Graph Neural Network (GNN) framework to address this gap. Our attention-based model predicts the number of upcoming clearances, the instructions issued to aircraft by ATCOs, from interactions within static traffic scenarios. Crucially, we derive an interpretable, per-aircraft task demand score by systematically ablating aircraft and measuring the impact on the model's predictions. Our framework significantly outperforms an ATCO-inspired heuristic and is a more reliable estimator of scenario complexity than established baselines. The resulting tool can attribute task demand to specific aircraft, offering a new way to analyse and understand the drivers of complexity for applications in controller training and airspace redesign.
Authors: Seyyed Saeid Cheshmi, Buyao Lyu, Thomas Lisko, Rajesh Rajamani, Robert A. McGovern, Yogatheesan Varatharajah
Abstract: Human Activity Recognition (HAR) based on wearable inertial sensors plays a critical role in remote health monitoring. In patients with movement disorders, the ability to detect abnormal patient movements in their home environments can enable continuous optimization of treatments and help alert caretakers as needed. Machine learning approaches have been proposed for HAR tasks using Inertial Measurement Unit (IMU) data; however, most rely on application-specific labels and lack generalizability to data collected in different environments or populations. To address this limitation, we propose a new cross-modal self-supervised pretraining approach to learn representations from large-sale unlabeled IMU-video data and demonstrate improved generalizability in HAR tasks on out of distribution (OOD) IMU datasets, including a dataset collected from patients with Parkinson's disease. Specifically, our results indicate that the proposed cross-modal pretraining approach outperforms the current state-of-the-art IMU-video pretraining approach and IMU-only pretraining under zero-shot and few-shot evaluations. Broadly, our study provides evidence that in highly dynamic data modalities, such as IMU signals, cross-modal pretraining may be a useful tool to learn generalizable data representations. Our software is available at https://github.com/scheshmi/IMU-Video-OOD-HAR.
Authors: Thomas Banker, Ali Mesbah
Abstract: Training sophisticated agents for optimal decision-making under uncertainty has been key to the rapid development of modern autonomous systems across fields. Notably, model-free reinforcement learning (RL) has enabled decision-making agents to improve their performance directly through system interactions, with minimal prior knowledge about the system. Yet, model-free RL has generally relied on agents equipped with deep neural network function approximators, appealing to the networks' expressivity to capture the agent's policy and value function for complex systems. However, neural networks amplify the issues of sample inefficiency, unsafe learning, and limited interpretability in model-free RL. To this end, this work introduces model-based agents as a compelling alternative for control policy approximation, leveraging adaptable models of system dynamics, cost, and constraints for safe policy learning. These models can encode prior system knowledge to inform, constrain, and aid in explaining the agent's decisions, while deficiencies due to model mismatch can be remedied with model-free RL. We outline the benefits and challenges of learning model-based agents -- exemplified by model predictive control -- and detail the primary learning approaches: Bayesian optimization, policy search RL, and offline strategies, along with their respective strengths. While model-free RL has long been established, its interplay with model-based agents remains largely unexplored, motivating our perspective on their combined potentials for sample-efficient learning of safe and interpretable decision-making agents.
Authors: Agata Kaczmarek (Warsaw University of Technology), Dawid P{\l}udowski (Warsaw University of Technology), Piotr Wilczy\'nski (Warsaw University of Technology), Przemys{\l}aw Biecek (Warsaw University of Technology), Krzysztof Kotowski (KP Labs), Ramez Shendy (KP Labs), Jakub Nalepa (KP Labs, Silesian University of Technology), Artur Janicki (Warsaw University of Technology), Evridiki Ntagiou (European Space Agency, European Space Operations Center)
Abstract: The "Fake or Real" competition hosted on Kaggle (\href{https://www.kaggle.com/competitions/fake-or-real-the-impostor-hunt}{https://www.kaggle.com/competitions/fake-or-real-the-impostor-hunt}) is the second part of a series of follow-up competitions and hackathons related to the "Assurance for Space Domain AI Applications" project funded by the European Space Agency (\href{https://assurance-ai.space-codev.org/}{https://assurance-ai.space-codev.org/}). The competition idea is based on two real-life AI security threats identified within the project -- data poisoning and overreliance in Large Language Models. The task is to distinguish between the proper output from LLM and the output generated under malicious modification of the LLM. As this problem was not extensively researched, participants are required to develop new techniques to address this issue or adjust already existing ones to this problem's statement.
URLs: https://www.kaggle.com/competitions/fake-or-real-the-impostor-hunt, https://www.kaggle.com/competitions/fake-or-real-the-impostor-hunt, https://assurance-ai.space-codev.org/, https://assurance-ai.space-codev.org/
Authors: Yongyi Yang, Hidenori Tanaka, Wei Hu
Abstract: In-context learning (ICL) enables large language models (LLMs) to acquire new behaviors from the input sequence alone without any parameter updates. Recent studies have shown that ICL can surpass the original meaning learned in pretraining stage through internalizing the structure the data-generating process (DGP) of the prompt into the hidden representations. However, the mechanisms by which LLMs achieve this ability is left open. In this paper, we present the first rigorous explanation of such phenomena by introducing a unified framework of double convergence, where hidden representations converge both over context and across layers. This double convergence process leads to an implicit bias towards smooth (low-frequency) representations, which we prove analytically and verify empirically. Our theory explains several open empirical observations, including why learned representations exhibit globally structured but locally distorted geometry, and why their total energy decays without vanishing. Moreover, our theory predicts that ICL has an intrinsic robustness towards high-frequency noise, which we empirically confirm. These results provide new insights into the underlying mechanisms of ICL, and a theoretical foundation to study it that hopefully extends to more general data distributions and settings.
Authors: Beka Begiashvili, Carlos J. Fernandez-Candel, Mat\'ias P\'erez Paredes
Abstract: Traditional echocardiographic parameters such as ejection fraction (EF) and global longitudinal strain (GLS) have limitations in the early detection of cardiac dysfunction. EF often remains normal despite underlying pathology, and GLS is influenced by load conditions and vendor variability. There is a growing need for reproducible, interpretable, and operator-independent parameters that capture subtle and global cardiac functional alterations. We introduce the Acoustic Index, a novel AI-derived echocardiographic parameter designed to quantify cardiac dysfunction from standard ultrasound views. The model combines Extended Dynamic Mode Decomposition (EDMD) based on Koopman operator theory with a hybrid neural network that incorporates clinical metadata. Spatiotemporal dynamics are extracted from echocardiographic sequences to identify coherent motion patterns. These are weighted via attention mechanisms and fused with clinical data using manifold learning, resulting in a continuous score from 0 (low risk) to 1 (high risk). In a prospective cohort of 736 patients, encompassing various cardiac pathologies and normal controls, the Acoustic Index achieved an area under the curve (AUC) of 0.89 in an independent test set. Cross-validation across five folds confirmed the robustness of the model, showing that both sensitivity and specificity exceeded 0.8 when evaluated on independent data. Threshold-based analysis demonstrated stable trade-offs between sensitivity and specificity, with optimal discrimination near this threshold. The Acoustic Index represents a physics-informed, interpretable AI biomarker for cardiac function. It shows promise as a scalable, vendor-independent tool for early detection, triage, and longitudinal monitoring. Future directions include external validation, longitudinal studies, and adaptation to disease-specific classifiers.
Authors: Rui Wang, Steven Klee, Alexis Roos
Abstract: This paper proposes using two metrics to quantify the forecastability of time series prior to model development: the spectral predictability score and the largest Lyapunov exponent. Unlike traditional model evaluation metrics, these measures assess the inherent forecastability characteristics of the data before any forecast attempts. The spectral predictability score evaluates the strength and regularity of frequency components in the time series, whereas the Lyapunov exponents quantify the chaos and stability of the system generating the data. We evaluated the effectiveness of these metrics on both synthetic and real-world time series from the M5 forecast competition dataset. Our results demonstrate that these two metrics can correctly reflect the inherent forecastability of a time series and have a strong correlation with the actual forecast performance of various models. By understanding the inherent forecastability of time series before model training, practitioners can focus their planning efforts on products and supply chain levels that are more forecastable, while setting appropriate expectations or seeking alternative strategies for products with limited forecastability.
Authors: Mrinal Mathur, Mike Doan, Barak Pearlmutter, Sergey Plis
Abstract: Transformers evaluated in a single, fixed-depth pass are provably limited in expressive power to the constant-depth circuit class TC0. Running a Transformer autoregressively removes that ceiling -- first in next-token prediction and, more recently, in chain-of-thought reasoning. Both regimes rely on feedback loops that decode internal states into tokens only to re-encode them in subsequent steps. While this "thinking aloud" mirrors human reasoning, biological brains iterate without externalising intermediate states as language. To boost the expressive power of encoder Transformers without resorting to token-level autoregression, we introduce the SELF-Transformer: an encoder layer that iteratively refines its own attention weights to a fixed point. Instead of producing -- in one pass -- the alignment matrix that remixes the input sequence, the SELF-Transformer iteratively updates that matrix internally, scaling test-time computation with input difficulty. This adaptivity yields up to 20\% accuracy gains on encoder-style benchmarks without increasing parameter count, demonstrating that input-adaptive alignment at test time offers substantial benefits for only a modest extra compute budget. Self-Transformers thus recover much of the expressive power of iterative reasoning while preserving the simplicity of pure encoder architectures.
Authors: Hanzhi Zhou (Taoyi), Erik Hornberger (Taoyi), Pengsheng Guo (Taoyi), Xiyou Zhou (Taoyi), Saiwen Wang (Taoyi), Xin Wang (Taoyi), Yifei He (Taoyi), Xuankai Chang (Taoyi), Rene Rauch (Taoyi), Louis D'hauwe (Taoyi), John Peebles (Taoyi), Alec Doane (Taoyi), Kohen Chia (Taoyi), Jenna Thibodeau (Taoyi), Zi-Yi Dou (Taoyi), Yuanyang Zhang (Taoyi), Ruoming Pang (Taoyi), Reed Li (Taoyi), Zhifeng Chen (Taoyi), Jeremy Warner (Taoyi), Zhaoyang Xu (Taoyi), Sophy Lee (Taoyi), David Mizrahi (Taoyi), Ramsey Tantawi (Taoyi), Chris Chaney (Taoyi), Kelsey Peterson (Taoyi), Jun Qin (Taoyi), Alex Dombrowski (Taoyi), Mira Chiang (Taoyi), Aiswarya Raghavan (Taoyi), Gerard Casamayor (Taoyi), Qibin Chen (Taoyi), Aonan Zhang (Taoyi), Nathalie Tran (Taoyi), Jianyu Wang (Taoyi), Hang Su (Taoyi), Thomas Voice (Taoyi), Alessandro Pappalardo (Taoyi), Brycen Wershing (Taoyi), Prasanth Yadla (Taoyi), Rui Li (Taoyi), Priyal Chhatrapati (Taoyi), Ismael Fernandez (Taoyi), Yusuf Goren (Taoyi), Xin Zheng (Taoyi), Forrest Huang (Taoyi), Tao Lei (Taoyi), Eray Yildiz (Taoyi), Alper Kokmen (Taoyi), Gokul Santhanam (Taoyi), Areeba Kamal (Taoyi), Kaan Elgin (Taoyi), Dian Ang Yap (Taoyi), Jeremy Liu (Taoyi), Peter Gray (Taoyi), Howard Xing (Taoyi), Kieran Liu (Taoyi), Matteo Ronchi (Taoyi), Moritz Schwarzer-Becker (Taoyi), Yun Zhu (Taoyi), Mandana Saebi (Taoyi), Jeremy Snow (Taoyi), David Griffiths (Taoyi), Guillaume Tartavel (Taoyi), Erin Feldman (Taoyi), Simon Lehnerer (Taoyi), Fernando Berm\'udez-Medina (Taoyi), Hans Han (Taoyi), Joe Zhou (Taoyi), Xiaoyi Ren (Taoyi), Sujeeth Reddy (Taoyi), Zirui Wang (Taoyi), Tom Gunter (Taoyi), Albert Antony (Taoyi), Yuanzhi Li (Taoyi), John Dennison (Taoyi), Tony Sun (Taoyi), Yena Han (Taoyi), Yi Qin (Taoyi), Sam Davarnia (Taoyi), Jeffrey Bigham (Taoyi), Wayne Shan (Taoyi), Hannah Gillis Coleman (Taoyi), Guillaume Klein (Taoyi), Peng Liu (Taoyi), Muyang Yu (Taoyi), Jack Cackler (Taoyi), Yuan Gao (Taoyi), Crystal Xiao (Taoyi), Binazir Karimzadeh (Taoyi), Zhengdong Zhang (Taoyi), Felix Bai (Taoyi), Albin Madappally Jose (Taoyi), Feng Nan (Taoyi), Nazir Kamaldin (Taoyi), Dong Yin (Taoyi), Hans Hao (Taoyi), Yanchao Sun (Taoyi), Yi Hua (Taoyi), Charles Maalouf (Taoyi), Alex Guillen Garcia (Taoyi), Guoli Yin (Taoyi), Lezhi Li (Taoyi), Mohana Prasad Sathya Moorthy (Taoyi), Hongbin Gao (Taoyi), Jay Tang (Taoyi), Joanna Arreaza-Taylor (Taoyi), Faye Lao (Taoyi), Carina Peng (Taoyi), Josh Shaffer (Taoyi), Dan Masi (Taoyi), Sushma Rao (Taoyi), Tommi Vehvilainen (Taoyi), Senyu Tong (Taoyi), Dongcai Shen (Taoyi), Yang Zhao (Taoyi), Chris Bartels (Taoyi), Peter Fu (Taoyi), Qingqing Cao (Taoyi), Christopher Neubauer (Taoyi), Ethan Li (Taoyi), Mingfei Gao (Taoyi), Rebecca Callahan (Taoyi), Richard Wei (Taoyi), Patrick Dong (Taoyi), Alex Braunstein (Taoyi), Sachin Ravi (Taoyi), Adolfo Lopez Mendez (Taoyi), Kaiwei Huang (Taoyi), Kun Duan (Taoyi), Haoshuo Huang (Taoyi), Rui Qian (Taoyi), Stefano Ligas (Taoyi), Jordan Huffaker (Taoyi), Dongxu Li (Taoyi), Bailin Wang (Taoyi), Nanzhu Wang (Taoyi), Anuva Agarwal (Taoyi), Tait Madsen (Taoyi), Josh Newnham (Taoyi), Abhishek Sharma (Taoyi), Zhile Ren (Taoyi), Deepak Gopinath (Taoyi), Erik Daxberger (Taoyi), Saptarshi Guha (Taoyi), Oron Levy (Taoyi), Jing Lu (Taoyi), Nan Dun (Taoyi), Marc Kirchner (Taoyi), Yinfei Yang (Taoyi), Manjot Bilkhu (Taoyi), Dave Nelson (Taoyi), Anthony Spalvieri-Kruse (Taoyi), Juan Lao Tebar (Taoyi), Yang Xu (Taoyi), Phani Mutyala (Taoyi), Gabriel Jacoby-Cooper (Taoyi), Yingbo Wang (Taoyi), Karla Vega (Taoyi), Vishaal Mahtani (Taoyi), Darren Botten (Taoyi), Eric Wang (Taoyi), Hanli Li (Taoyi), Matthias Paulik (Taoyi), Haoran Yan (Taoyi), Navid Shiee (Taoyi), Yihao Qian (Taoyi), Bugu Wu (Taoyi), Qi Zhu (Taoyi), Ob Adaranijo (Taoyi), Bhuwan Dhingra (Taoyi), Zhe Gan (Taoyi), Nicholas Seidl (Taoyi), Grace Duanmu (Taoyi), Rong Situ (Taoyi), Yiping Ma (Taoyi), Yin Xia (Taoyi), David Riazati (Taoyi), Vasileios Saveris (Taoyi), Anh Nguyen (Taoyi), Michael (Taoyi), Lee, Patrick Sonnenberg, Chinguun Erdenebileg, Yanghao Li, Vivian Ma, James Chou, Isha Garg, Mark Lee, Keen You, Yuhong Li, Ransen Niu, Nandhitha Raghuram, Pulkit Agrawal, Henry Mason, Sumeet Singh, Keyu He, Hong-You Chen, Lucas Guibert, Shiyu Li, Varsha Paidi, Narendran Raghavan, Mingze Xu, Yuli Yang, Sergiu Sima, Irina Belousova, Sprite Chu, Afshin Dehghan, Philipp Dufter, David Haldimann, Zhen Yang, Margit Bowler, Chang Liu, Ying-Chang Cheng, Vivek Rathod, Syd Evans, Wilson Tsao, Dustin Withers, Haitian Sun, Biyao Wang, Peter Grasch, Walker Cheng, Yihao Feng, Vivek Kumar, Frank Chu, Victoria M\"onchJuan Haladjian, Doug Kang, Jiarui Lu, Ciro Sannino, Max Lam, Floris Weers, Bowen Pan, Kenneth Jung, Dhaval Doshi, Fangping Shi, Olli Saarikivi, Alp Aygar, Josh Elman, Cheng Leong, Eshan Verma, Matthew Lei, Jeff Nichols, Jiulong Shan, Donald Zhang, Lawrence Zhou, Stephen Murphy, Xianzhi Du, Chang Lan, Ankur Jain, Elmira Amirloo, Marcin Eichner, Naomy Sabo, Anupama Mann Anupama, David Qiu, Zhao Meng, Michael FitzMaurice, Peng Zhang, Simon Yeung, Chen Chen, Marco Zuliani, Andrew Hansen, Yang Lu, Brent Ramerth, Ziyi Zhong, Parsa Mazaheri, Matthew Hopkins, Mengyu Li, Simon Wang, David Chen, Farzin Rasteh, Chong Wang, Josh Gardner, Asaf Liberman, Haoxuan You, Andrew Walkingshaw, Xingyu Zhou, Jinhao Lei, Yan Meng, Quentin Keunebroek, Sam Wiseman, Anders Boesen Lindbo Larsen, Yi Zhang, Zaid Ahmed, Haiming Gang, Aaron Franklin, Kelvin Zou, Guillaume Seguin, Jonathan Janke, Rachel Burger, Co Giang, Cheng Shen, Jen Liu, Sanskruti Shah, Xiang Kong, Yiran Fei, TJ Collins, Chen Zhang, Zhiyun Lu, Michael Booker, Qin Ba, Yasutaka Tanaka, Andres Romero Mier Y Teran, Federico Scozzafava, Regan Poston, Jane Li, Eduardo Jimenez, Bas Straathof, Karanjeet Singh, Lindsay Hislop, Rajat Arora, Deepa Seshadri, Boyue Li, Colorado Reed, Zhen Li, TJ Lu, Yi Wang, Kaelen Haag, Nicholas Lusskin, Raunak Sinha, Rahul Nair, Eldon Schoop, Mary Beth Kery, Mehrdad Farajtbar, Brenda Yang, George Horrell, Shiwen Zhao, Dhruti Shah, Cha Chen, Bowen Zhang, Chang Gao, Devi Krishna, Jennifer Mallalieu, Javier Movellan, Di Feng, Emily Zhang, Sam Xu, Junting Pan, Dominik Moritz, Suma Jayaram, Kevin Smith, Dongseong Hwang, Daniel Parilla, Jiaming Hu, You-Cyuan Jhang, Emad Soroush, Fred Hohman, Nan Du, Emma Wang, Sam Dodge, Pragnya Sridhar, Joris Pelemans, Wei Fang, Nina Wenzel, Joseph Yitan Cheng, Hadas Kotek, Chung-Cheng Chiu, Meng Cao, Haijing Fu, Ruixuan Hou, Ke Ye, Diane Zhu, Nikhil Bhendawade, Joseph Astrauskas, Jian Liu, Sai Aitharaju, Wentao Wu, Artsiom Peshko, Hyunjik Kim, Nilesh Shahdadpuri, Andy De Wang, Qi Shan, Piotr Maj, Raul Rea Menacho, Justin Lazarow, Eric Liang Yang, Arsalan Farooq, Donghan Yu, David G\"uera, Minsik Cho, Kavya Nerella, Yongqiang Wang, Tao Jia, John Park, Jeff Lai, Haotian Zhang, Futang Peng, Daniele Molinari, Aparna Rajamani, Tyler Johnson, Lauren Gardiner, Chao Jia, Violet Yao, Wojciech Kryscinski, Xiujun Li, Shang-Chen Wu
Abstract: We introduce two multilingual, multimodal foundation language models that power Apple Intelligence features across Apple devices and services: i a 3B-parameter on-device model optimized for Apple silicon through architectural innovations such as KV-cache sharing and 2-bit quantization-aware training; and ii a scalable server model built on a novel Parallel-Track Mixture-of-Experts PT-MoE transformer that combines track parallelism, mixture-of-experts sparse computation, and interleaved global-local attention to deliver high quality with competitive cost on Apple's Private Cloud Compute platform. Both models are trained on large-scale multilingual and multimodal datasets sourced via responsible web crawling, licensed corpora, and high-quality synthetic data, then further refined with supervised fine-tuning and reinforcement learning on a new asynchronous platform. The resulting models support several additional languages while understanding images and executing tool calls. In public benchmarks and human evaluations, both the server model and the on-device model match or surpass comparably sized open baselines. A new Swift-centric Foundation Models framework exposes guided generation, constrained tool calling, and LoRA adapter fine-tuning, allowing developers to integrate these capabilities with a few lines of code. The latest advancements in Apple Intelligence models are grounded in our Responsible AI approach with safeguards like content filtering and locale-specific evaluation, as well as our commitment to protecting our users' privacy with innovations like Private Cloud Compute.
Authors: Hyunji Nam, Yanming Wan, Mickel Liu, Jianxun Lian, Natasha Jaques
Abstract: As everyday use cases of large language model (LLM) AI assistants have expanded, it is becoming increasingly important to personalize responses to align to different users' preferences and goals. While reinforcement learning from human feedback (RLHF) is effective at improving LLMs to be generally more helpful and fluent, it does not account for variability across users, as it models the entire user population with a single reward model. We present a novel framework, Preference Learning Using Summarization (PLUS), that learns text-based summaries of each user's preferences, characteristics, and past conversations. These summaries condition the reward model, enabling it to make personalized predictions about the types of responses valued by each user. We train the user-summarization model with reinforcement learning, and update the reward model simultaneously, creating an online co-adaptation loop. We show that in contrast with prior personalized RLHF techniques or with in-context learning of user information, summaries produced by PLUS capture meaningful aspects of a user's preferences. Across different pluralistic user datasets, we show that our method is robust to new users and diverse conversation topics. Additionally, we demonstrate that the textual summaries generated about users can be transferred for zero-shot personalization of stronger, proprietary models like GPT-4. The resulting user summaries are not only concise and portable, they are easy for users to interpret and modify, allowing for more transparency and user control in LLM alignment.
Authors: Yudai Hayashi, Shuhei Goda, Yuta Saito
Abstract: Matching users based on mutual preferences is a fundamental aspect of services driven by reciprocal recommendations, such as job search and dating applications. Although A/B tests remain the gold standard for evaluating new policies in recommender systems for matching markets, it is costly and impractical for frequent policy updates. Off-Policy Evaluation (OPE) thus plays a crucial role by enabling the evaluation of recommendation policies using only offline logged data naturally collected on the platform. However, unlike conventional recommendation settings, the large scale and bidirectional nature of user interactions in matching platforms introduce variance issues and exacerbate reward sparsity, making standard OPE methods unreliable. To address these challenges and facilitate effective offline evaluation, we propose novel OPE estimators, \textit{DiPS} and \textit{DPR}, specifically designed for matching markets. Our methods combine elements of the Direct Method (DM), Inverse Propensity Score (IPS), and Doubly Robust (DR) estimators while incorporating intermediate labels, such as initial engagement signals, to achieve better bias-variance control in matching markets. Theoretically, we derive the bias and variance of the proposed estimators and demonstrate their advantages over conventional methods. Furthermore, we show that these estimators can be seamlessly extended to offline policy learning methods for improving recommendation policies for making more matches. We empirically evaluate our methods through experiments on both synthetic data and A/B testing logs from a real job-matching platform. The empirical results highlight the superiority of our approach over existing methods in off-policy evaluation and learning tasks for a variety of configurations.
Authors: Binxiong Li, Yuefei Wang, Xu Xiang, Xue Li, Binyu Zhao, Heyang Gao, Qinyu Zhao, Xi Yu
Abstract: In recent years, models based on Graph Convolutional Networks (GCN) have made significant strides in the field of graph data analysis. However, challenges such as over-smoothing and over-compression remain when handling large-scale and complex graph datasets, leading to a decline in clustering quality. Although the Graph Transformer architecture has mitigated some of these issues, its performance is still limited when processing heterogeneous graph data. To address these challenges, this study proposes a novel deep clustering framework that comprising GCN, Autoencoder (AE), and Graph Transformer, termed the Tri-Learn Graph Fusion Network (Tri-GFN). This framework enhances the differentiation and consistency of global and local information through a unique tri-learning mechanism and feature fusion enhancement strategy. The framework integrates GCN, AE, and Graph Transformer modules. These components are meticulously fused by a triple-channel enhancement module, which maximizes the use of both node attributes and topological structures, ensuring robust clustering representation. The tri-learning mechanism allows mutual learning among these modules, while the feature fusion strategy enables the model to capture complex relationships, yielding highly discriminative representations for graph clustering. It surpasses many state-of-the-art methods, achieving an accuracy improvement of approximately 0.87% on the ACM dataset, 14.14 % on the Reuters dataset, and 7.58 % on the USPS dataset. Due to its outstanding performance on the Reuters dataset, Tri-GFN can be applied to automatic news classification, topic retrieval, and related fields.
Authors: Daniel Commey, Kamel Abbad, Garth V. Crosby, Lyes Khoukhi
Abstract: Communication overhead remains a primary bottleneck in federated learning (FL), particularly for applications involving mobile and IoT devices with constrained bandwidth. This work introduces FedSkipTwin, a novel client-skipping algorithm driven by lightweight, server-side digital twins. Each twin, implemented as a simple LSTM, observes a client's historical sequence of gradient norms to forecast both the magnitude and the epistemic uncertainty of its next update. The server leverages these predictions, requesting communication only when either value exceeds a predefined threshold; otherwise, it instructs the client to skip the round, thereby saving bandwidth. Experiments are conducted on the UCI-HAR and MNIST datasets with 10 clients under a non-IID data distribution. The results demonstrate that FedSkipTwin reduces total communication by 12-15.5% across 20 rounds while simultaneously improving final model accuracy by up to 0.5 percentage points compared to the standard FedAvg algorithm. These findings establish that prediction-guided skipping is a practical and effective strategy for resource-aware FL in bandwidth-constrained edge environments.
Authors: Nimisha Ghosh, Daniele Santoni, Debaleena Nawn, Eleonora Ottaviani, Giovanni Felici
Abstract: The impact of Transformer-based language models has been unprecedented in Natural Language Processing (NLP). The success of such models has also led to their adoption in other fields including bioinformatics. Taking this into account, this paper discusses recent advances in Transformer-based models for protein sequence analysis and design. In this review, we have discussed and analysed a significant number of works pertaining to such applications. These applications encompass gene ontology, functional and structural protein identification, generation of de novo proteins and binding of proteins. We attempt to shed light on the strength and weaknesses of the discussed works to provide a comprehensive insight to readers. Finally, we highlight shortcomings in existing research and explore potential avenues for future developments. We believe that this review will help researchers working in this field to have an overall idea of the state of the art in this field, and to orient their future studies.
Authors: Yue Yang, Zihan Su, Ying Zhang, Chang Chuan Goh, Yuxiang Lin, Anthony Graham Bellotti, Boon Giin Lee
Abstract: This study addresses a critical challenge in time series anomaly detection: enhancing the predictive capability of loan default models more than three months in advance to enable early identification of default events, helping financial institutions implement preventive measures before risk events materialize. Existing methods have significant drawbacks, such as their lack of accuracy in early predictions and their dependence on training and testing within the same year and specific time frames. These issues limit their practical use, particularly with out-of-time data. To address these, the study introduces two innovative architectures, GRU-KAN and LSTM-KAN, which merge Kolmogorov-Arnold Networks (KAN) with Gated Recurrent Units (GRU) and Long Short-Term Memory (LSTM) networks. The proposed models were evaluated against the baseline models (LSTM, GRU, LSTM-Attention, and LSTM-Transformer) in terms of accuracy, precision, recall, F1 and AUC in different lengths of feature window, sample sizes, and early prediction intervals. The results demonstrate that the proposed model achieves a prediction accuracy of over 92% three months in advance and over 88% eight months in advance, significantly outperforming existing baselines.
Authors: Martin Krutsk\'y, Gustav \v{S}\'ir, Vyacheslav Kungurtsev, Georgios Korpas
Abstract: Physics-inspired graph neural networks (PI-GNNs) have been utilized as an efficient unsupervised framework for relaxing combinatorial optimization problems encoded through a specific graph structure and loss, reflecting dependencies between the problem's variables. While the framework has yielded promising results in various combinatorial problems, we show that the performance of PI-GNNs systematically plummets with an increasing density of the combinatorial problem graphs. Our analysis reveals an interesting phase transition in the PI-GNNs' training dynamics, associated with degenerate solutions for the denser problems, highlighting a discrepancy between the relaxed, real-valued model outputs and the binary-valued problem solutions. To address the discrepancy, we propose principled alternatives to the naive strategy used in PI-GNNs by building on insights from fuzzy logic and binarized neural networks. Our experiments demonstrate that the portfolio of proposed methods significantly improves the performance of PI-GNNs in increasingly dense settings.
Authors: Anabel Yong, Austin Tripp, Layla Hosseini-Gerami, Brooks Paige
Abstract: Multi-objective Bayesian optimization (MOBO) provides a principled framework for navigating trade-offs in molecular design. However, its empirical advantages over scalarized alternatives remain underexplored. We benchmark a simple Pareto-based MOBO strategy -- Expected Hypervolume Improvement (EHVI) -- against a simple fixed-weight scalarized baseline using Expected Improvement (EI), under a tightly controlled setup with identical Gaussian Process surrogates and molecular representations. Across three molecular optimization tasks, EHVI consistently outperforms scalarized EI in terms of Pareto front coverage, convergence speed, and chemical diversity. While scalarization encompasses flexible variants -- including random or adaptive schemes -- our results show that even strong deterministic instantiations can underperform in low-data regimes. These findings offer concrete evidence for the practical advantages of Pareto-aware acquisition in de novo molecular optimization, especially when evaluation budgets are limited and trade-offs are nontrivial.
Authors: Hao Wang, Yu Liu, Daniel Biggs, Haoru Wang, Jiandong Yu, Ping Huang
Abstract: Simulating interactions between deformable bodies is vital in fields like material science, mechanical design, and robotics. While learning-based methods with Graph Neural Networks (GNNs) are effective at solving complex physical systems, they encounter scalability issues when modeling deformable body interactions. To model interactions between objects, pairwise global edges have to be created dynamically, which is computationally intensive and impractical for large-scale meshes. To overcome these challenges, drawing on insights from geometric representations, we propose an Adaptive Spatial Tokenization (AST) method for efficient representation of physical states. By dividing the simulation space into a grid of cells and mapping unstructured meshes onto this structured grid, our approach naturally groups adjacent mesh nodes. We then apply a cross-attention module to map the sparse cells into a compact, fixed-length embedding, serving as tokens for the entire physical state. Self-attention modules are employed to predict the next state over these tokens in latent space. This framework leverages the efficiency of tokenization and the expressive power of attention mechanisms to achieve accurate and scalable simulation results. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art approaches in modeling deformable body interactions. Notably, it remains effective on large-scale simulations with meshes exceeding 100,000 nodes, where existing methods are hindered by computational limitations. Additionally, we contribute a novel large-scale dataset encompassing a wide range of deformable body interactions to support future research in this area.
Authors: Danilo Avola, Andrea Bernardini, Giancarlo Crocetti, Andrea Ladogana, Mario Lezoche, Maurizio Mancini, Daniele Pannone, Amedeo Ranaldi
Abstract: Parkinson's Disease PD is a progressive neurodegenerative disorder that affects motor and cognitive functions with early diagnosis being critical for effective clinical intervention Electroencephalography EEG offers a noninvasive and costeffective means of detecting PDrelated neural alterations yet the development of reliable automated diagnostic models remains a challenge In this study we conduct a systematic benchmark of traditional machine learning ML and deep learning DL models for classifying PD using a publicly available oddball task dataset Our aim is to lay the groundwork for developing an effective learning system and to determine which approach produces the best results We implement a unified sevenstep preprocessing pipeline and apply consistent subjectwise crossvalidation and evaluation criteria to ensure comparability across models Our results demonstrate that while baseline deep learning architectures particularly CNNLSTM models achieve the best performance compared to other deep learning architectures underlining the importance of capturing longrange temporal dependencies several traditional classifiers such as XGBoost also offer strong predictive accuracy and calibrated decision boundaries By rigorously comparing these baselines our work provides a solid reference framework for future studies aiming to develop and evaluate more complex or specialized architectures Establishing a reliable set of baseline results is essential to contextualize improvements introduced by novel methods ensuring scientific rigor and reproducibility in the evolving field of EEGbased neurodiagnostics
Authors: Danilo Avola, Muhammad Yasir Bilal, Emad Emam, Cristina Lakasz, Daniele Pannone, Amedeo Ranaldi
Abstract: Deception detection is a significant challenge in fields such as security, psychology, and forensics. This study presents a deep learning approach for classifying deceptive and truthful behavior using ElectroEncephaloGram (EEG) signals from the Bag-of-Lies dataset, a multimodal corpus designed for naturalistic, casual deception scenarios. A Bidirectional Gated Recurrent Unit (Bi-GRU) neural network was trained to perform binary classification of EEG samples. The model achieved a test accuracy of 97\%, along with high precision, recall, and F1-scores across both classes. These results demonstrate the effectiveness of using bidirectional temporal modeling for EEG-based deception detection and suggest potential for real-time applications and future exploration of advanced neural architectures.
Authors: Zizhao Zhang, Tianxiang Zhao, Yu Sun, Liping Sun, Jichuan Kang
Abstract: To address the challenges posed by cascading reactions caused by component failures in autonomous cargo ships (ACS) and the uncertainties in emergency decision-making, this paper proposes a novel hybrid feature fusion framework for constructing a graph-structured dataset of failure modes. By employing an improved cuckoo search algorithm (HN-CSA), the literature retrieval efficiency is significantly enhanced, achieving improvements of 7.1% and 3.4% compared to the NSGA-II and CSA search algorithms, respectively. A hierarchical feature fusion framework is constructed, using Word2Vec encoding to encode subsystem/component features, BERT-KPCA to process failure modes/reasons, and Sentence-BERT to quantify the semantic association between failure impact and emergency decision-making. The dataset covers 12 systems, 1,262 failure modes, and 6,150 propagation paths. Validation results show that the GATE-GNN model achieves a classification accuracy of 0.735, comparable to existing benchmarks. Additionally, a silhouette coefficient of 0.641 indicates that the features are highly distinguishable. In the label prediction results, the Shore-based Meteorological Service System achieved an F1 score of 0.93, demonstrating high prediction accuracy. This paper not only provides a solid foundation for failure analysis in autonomous cargo ships but also offers reliable support for fault diagnosis, risk assessment, and intelligent decision-making systems. The link to the dataset is https://github.com/wojiufukele/Graph-Structured-about-CSA.
URLs: https://github.com/wojiufukele/Graph-Structured-about-CSA.
Authors: Ren\'e Heinrich, Lukas Rauch, Bernhard Sick, Christoph Scholz
Abstract: Adversarial training is a promising strategy for enhancing model robustness against adversarial attacks. However, its impact on generalization under substantial data distribution shifts in audio classification remains largely unexplored. To address this gap, this work investigates how different adversarial training strategies improve generalization performance and adversarial robustness in audio classification. The study focuses on two model architectures: a conventional convolutional neural network (ConvNeXt) and an inherently interpretable prototype-based model (AudioProtoPNet). The approach is evaluated using a challenging bird sound classification benchmark. This benchmark is characterized by pronounced distribution shifts between training and test data due to varying environmental conditions and recording methods, a common real-world challenge. The investigation explores two adversarial training strategies: one based on output-space attacks that maximize the classification loss function, and another based on embedding-space attacks designed to maximize embedding dissimilarity. These attack types are also used for robustness evaluation. Additionally, for AudioProtoPNet, the study assesses the stability of its learned prototypes under targeted embedding-space attacks. Results show that adversarial training, particularly using output-space attacks, improves clean test data performance by an average of 10.5% relative and simultaneously strengthens the adversarial robustness of the models. These findings, although derived from the bird sound domain, suggest that adversarial training holds potential to enhance robustness against both strong distribution shifts and adversarial attacks in challenging audio classification settings.
Authors: Matthias Jobst, Tim Langer, Chen Liu, Mehmet Alici, Hector A. Gonzalez, Christian Mayr
Abstract: This work presents a multi-layer DNN scheduling framework as an extension of OctopuScheduler, providing an end-to-end flow from PyTorch models to inference on a single SpiNNaker2 chip. Together with a front-end comprised of quantization and lowering steps, the proposed framework enables the edge-based execution of large and complex DNNs up to transformer scale using the neuromorphic platform SpiNNaker2.
Authors: Shangyou Wang, Zezhong Ding, Xike Xie
Abstract: Graph Neural Networks (GNNs) have shown remarkable success in graph classification tasks by capturing both structural and feature-based representations. However, real-world graphs often exhibit two critical forms of imbalance: class imbalance and graph size imbalance. These imbalances can bias the learning process and degrade model performance. Existing methods typically address only one type of imbalance or incur high computational costs. In this work, we propose SamGoG, a sampling-based Graph-of-Graphs (GoG) learning framework that effectively mitigates both class and graph size imbalance. SamGoG constructs multiple GoGs through an efficient importance-based sampling mechanism and trains on them sequentially. This sampling mechanism incorporates the learnable pairwise similarity and adaptive GoG node degree to enhance edge homophily, thus improving downstream model quality. SamGoG can seamlessly integrate with various downstream GNNs, enabling their efficient adaptation for graph classification tasks. Extensive experiments on benchmark datasets demonstrate that SamGoG achieves state-of-the-art performance with up to a 15.66% accuracy improvement with 6.7$\times$ training acceleration.
Authors: Oussama Bouaggad, Natalia Grabar
Abstract: In the fast-moving world of AI, as organizations and researchers develop more advanced models, they face challenges due to their sheer size and computational demands. Deploying such models on edge devices or in resource-constrained environments adds further challenges related to energy consumption, memory usage and latency. To address these challenges, emerging trends are shaping the future of efficient model optimization techniques. From this premise, by employing supervised state-of-the-art transformer-based models, this research introduces a systematic method for ontology alignment, grounded in cosine-based semantic similarity between a biomedical layman vocabulary and the Unified Medical Language System (UMLS) Metathesaurus. It leverages Microsoft Olive to search for target optimizations among different Execution Providers (EPs) using the ONNX Runtime backend, followed by an assembled process of dynamic quantization employing Intel Neural Compressor and IPEX (Intel Extension for PyTorch). Through our optimization process, we conduct extensive assessments on the two tasks from the DEFT 2020 Evaluation Campaign, achieving a new state-of-the-art in both. We retain performance metrics intact, while attaining an average inference speed-up of 20x and reducing memory usage by approximately 70%.
Authors: Yaowei Jin, Junjie Wang, Wenkai Xiang, Duanhua Cao, Dan Teng, Zhehuan Fan, Jiacheng Xiong, Xia Sheng, Chuanlong Zeng, Mingyue Zheng, Qian Shi
Abstract: Advances in deep learning for molecular generation show promise in accelerating drug discovery. Bayesian Flow Networks (BFNs) have recently shown impressive performance across diverse chemical tasks, with their success often ascribed to the paradigm of modeling in a low-variance parameter space. However, the Bayesian inference-based strategy imposes limitations on designing more flexible distribution transformation pathways, making it challenging to adapt to diverse data distributions and varied task requirements. Furthermore, the potential for simpler, more efficient parameter-space-based models is unexplored. To address this, we propose a novel Parameter Interpolation Flow model (named PIF) with detailed theoretical foundation, training, and inference procedures. We then develop MolPIF for structure-based drug design, demonstrating its superior performance across diverse metrics compared to baselines. This work validates the effectiveness of parameter-space-based generative modeling paradigm for molecules and offers new perspectives for model design.
Authors: Enhao Cheng, Shoujia Zhang, Jianhua Yin, Li Jin, Liqiang Nie
Abstract: Graph clustering is crucial for unraveling intricate data structures, yet it presents significant challenges due to its unsupervised nature. Recently, goal-directed clustering techniques have yielded impressive results, with contrastive learning methods leveraging pseudo-label garnering considerable attention. Nonetheless, pseudo-label as a supervision signal is unreliable and existing goal-directed approaches utilize only features to construct a single-target distribution for single-center optimization, which lead to incomplete and less dependable guidance. In our work, we propose a novel Dual-Center Graph Clustering (DCGC) approach based on neighbor distribution properties, which includes representation learning with neighbor distribution and dual-center optimization. Specifically, we utilize neighbor distribution as a supervision signal to mine hard negative samples in contrastive learning, which is reliable and enhances the effectiveness of representation learning. Furthermore, neighbor distribution center is introduced alongside feature center to jointly construct a dual-target distribution for dual-center optimization. Extensive experiments and analysis demonstrate superior performance and effectiveness of our proposed method.
Authors: Tim Rensmeyer, Denis Kramer, Oliver Niggemann
Abstract: Due to the computational complexity of evaluating interatomic forces from first principles, the creation of interatomic machine learning force fields has become a highly active field of research. However, the generation of training datasets of sufficient size and sample diversity itself comes with a computational burden that can make this approach impractical for modeling rare events or systems with a large configuration space. Fine-tuning foundation models that have been pre-trained on large-scale material or molecular databases offers a promising opportunity to reduce the amount of training data necessary to reach a desired level of accuracy. However, even if this approach requires less training data overall, creating a suitable training dataset can still be a very challenging problem, especially for systems with rare events and for end-users who don't have an extensive background in machine learning. In on-the-fly learning, the creation of a training dataset can be largely automated by using model uncertainty during the simulation to decide if the model is accurate enough or if a structure should be recalculated with classical methods and used to update the model. A key challenge for applying this form of active learning to the fine-tuning of foundation models is how to assess the uncertainty of those models during the fine-tuning process, even though most foundation models lack any form of uncertainty quantification. In this paper, we overcome this challenge by introducing a fine-tuning approach based on Bayesian neural network methods and a subsequent on-the-fly workflow that automatically fine-tunes the model while maintaining a pre-specified accuracy and can detect rare events such as transition states and sample them at an increased rate relative to their occurrence.
Authors: Aditi Anand, Suman Banerjee, Dildar Ali
Abstract: In Reinforcement Learning (abbreviated as RL), an agent interacts with the environment via a set of possible actions, and a reward is generated from some unknown distribution. The task here is to find an optimal set of actions such that the reward after a certain time step gets maximized. In a traditional setup, the reward function in an RL Problem is considered additive. However, in reality, there exist many problems, including path planning, coverage control, etc., the reward function follows the diminishing return, which can be modeled as a submodular function. In this paper, we study a variant of the RL Problem where the reward function is submodular, and our objective is to find an optimal policy such that this reward function gets maximized. We have proposed a pruned submodularity graph-based approach that provides a provably approximate solution in a feasible computation time. The proposed approach has been analyzed to understand its time and space requirements as well as a performance guarantee. We have experimented with a benchmark agent-environment setup, which has been used for similar previous studies, and the results are reported. From the results, we observe that the policy obtained by our proposed approach leads to more reward than the baseline methods.
Authors: Kevin Dradjat, Massinissa Hamidi, Pierre Bartet, Blaise Hanczar
Abstract: Predicting phenotypes from gene expression data is a crucial task in biomedical research, enabling insights into disease mechanisms, drug responses, and personalized medicine. Traditional machine learning and deep learning rely on supervised learning, which requires large quantities of labeled data that are costly and time-consuming to obtain in the case of gene expression data. Self-supervised learning has recently emerged as a promising approach to overcome these limitations by extracting information directly from the structure of unlabeled data. In this study, we investigate the application of state-of-the-art self-supervised learning methods to bulk gene expression data for phenotype prediction. We selected three self-supervised methods, based on different approaches, to assess their ability to exploit the inherent structure of the data and to generate qualitative representations which can be used for downstream predictive tasks. By using several publicly available gene expression datasets, we demonstrate how the selected methods can effectively capture complex information and improve phenotype prediction accuracy. The results obtained show that self-supervised learning methods can outperform traditional supervised models besides offering significant advantage by reducing the dependency on annotated data. We provide a comprehensive analysis of the performance of each method by highlighting their strengths and limitations. We also provide recommendations for using these methods depending on the case under study. Finally, we outline future research directions to enhance the application of self-supervised learning in the field of gene expression data analysis. This study is the first work that deals with bulk RNA-Seq data and self-supervised learning.
Authors: Turan Orujlu, Christian Gumbsch, Martin V. Butz, Charley M Wu
Abstract: Formal frameworks of causality have operated largely parallel to modern trends in deep reinforcement learning (RL). However, there has been a revival of interest in formally grounding the representations learned by neural networks in causal concepts. Yet, most attempts at neural models of causality assume static causal graphs and ignore the dynamic nature of causal interactions. In this work, we introduce Causal Process framework as a novel theory for representing dynamic hypotheses about causal structure. Furthermore, we present Causal Process Model as an implementation of this framework. This allows us to reformulate the attention mechanism popularized by Transformer networks within an RL setting with the goal to infer interpretable causal processes from visual observations. Here, causal inference corresponds to constructing a causal graph hypothesis which itself becomes an RL task nested within the original RL problem. To create an instance of such hypothesis, we employ RL agents. These agents establish links between units similar to the original Transformer attention mechanism. We demonstrate the effectiveness of our approach in an RL environment where we outperform current alternatives in causal representation learning and agent performance, and uniquely recover graphs of dynamic causal processes.
Authors: Jingbo Liang, Bruna Jacobson
Abstract: Extensively exploring protein conformational landscapes remains a major challenge in computational biology due to the high computational cost involved in dynamic physics-based simulations. In this work, we propose a novel pipeline, MoDyGAN, that leverages molecular dynamics (MD) simulations and generative adversarial networks (GANs) to explore protein conformational spaces. MoDyGAN contains a generator that maps Gaussian distributions into MD-derived protein trajectories, and a refinement module that combines ensemble learning with a dual-discriminator to further improve the plausibility of generated conformations. Central to our approach is an innovative representation technique that reversibly transforms 3D protein structures into 2D matrices, enabling the use of advanced image-based GAN architectures. We use three rigid proteins to demonstrate that MoDyGAN can generate plausible new conformations. We also use deca-alanine as a case study to show that interpolations within the latent space closely align with trajectories obtained from steered molecular dynamics (SMD) simulations. Our results suggest that representing proteins as image-like data unlocks new possibilities for applying advanced deep learning techniques to biomolecular simulation, leading to an efficient sampling of conformational states. Additionally, the proposed framework holds strong potential for extension to other complex 3D structures.
Authors: Yifan Wei, Anwar Said, Waseem Abbas, Xenofon Koutsoukos
Abstract: Anomaly detection in complex domains poses significant challenges due to the need for extensive labeled data and the inherently imbalanced nature of anomalous versus benign samples. Graph-based machine learning models have emerged as a promising solution that combines attribute and relational data to uncover intricate patterns. However, the scarcity of anomalous data exacerbates the challenge, which requires innovative strategies to enhance model learning with limited information. In this paper, we hypothesize that the incorporation of the influence of the nodes, quantified through average controllability, can significantly improve the performance of anomaly detection. We propose two novel approaches to integrate average controllability into graph-based frameworks: (1) using average controllability as an edge weight and (2) encoding it as a one-hot edge attribute vector. Through rigorous evaluation on real-world and synthetic networks with six state-of-the-art baselines, our proposed methods demonstrate improved performance in identifying anomalies, highlighting the critical role of controllability measures in enhancing the performance of graph machine learning models. This work underscores the potential of integrating average controllability as additional metrics to address the challenges of anomaly detection in sparse and imbalanced datasets.
Authors: Eli Verwimp, Gustav Ryberg Smidt, Hendrik Hameeuw, Katrien De Graef
Abstract: The work in this paper describes the training and evaluation of machine learning (ML) techniques for the classification of cuneiform signs. There is a lot of variability in cuneiform signs, depending on where they come from, for what and by whom they were written, but also how they were digitized. This variability makes it unlikely that an ML model trained on one dataset will perform successfully on another dataset. This contribution studies how such differences impact that performance. Based on our results and insights, we aim to influence future data acquisition standards and provide a solid foundation for future cuneiform sign classification tasks. The ML model has been trained and tested on handwritten Old Babylonian (c. 2000-1600 B.C.E.) documentary texts inscribed on clay tablets originating from three Mesopotamian cities (Nippur, D\=ur-Abie\v{s}uh and Sippar). The presented and analysed model is ResNet50, which achieves a top-1 score of 87.1% and a top-5 score of 96.5% for signs with at least 20 instances. As these automatic classification results are the first on Old Babylonian texts, there are currently no comparable results.
Authors: Jagruti Patel, Thomas A. W. Bolton, Mikkel Sch\"ottner, Anjali Tarun, Sebastien Tourbier, Yasser Alem\`an-G\`omez, Jonas Richiardi, Patric Hagmann
Abstract: Small sample sizes in neuroimaging in general, and in structural connectome (SC) studies in particular limit the development of reliable biomarkers for neurological and psychiatric disorders - such as Alzheimer's disease and schizophrenia - by reducing statistical power, reliability, and generalizability. Large-scale multi-site studies have exist, but they have acquisition-related biases due to scanner heterogeneity, compromising imaging consistency and downstream analyses. While existing SC harmonization methods - such as linear regression (LR), ComBat, and deep learning techniques - mitigate these biases, they often rely on detailed metadata, traveling subjects (TS), or overlook the graph-topology of SCs. To address these limitations, we propose a site-conditioned deep harmonization framework that harmonizes SCs across diverse acquisition sites without requiring metadata or TS that we test in a simulated scenario based on the Human Connectome Dataset. Within this framework, we benchmark three deep architectures - a fully connected autoencoder (AE), a convolutional AE, and a graph convolutional AE - against a top-performing LR baseline. While non-graph models excel in edge-weight prediction and edge existence detection, the graph AE demonstrates superior preservation of topological structure and subject-level individuality, as reflected by graph metrics and fingerprinting accuracy, respectively. Although the LR baseline achieves the highest numerical performance by explicitly modeling acquisition parameters, it lacks applicability to real-world multi-site use cases as detailed acquisition metadata is often unavailable. Our results highlight the critical role of model architecture in SC harmonization performance and demonstrate that graph-based approaches are particularly well-suited for structure-aware, domain-generalizable SC harmonization in large-scale multi-site SC studies.
Authors: Itay Katav, Aryeh Kontorovich
Abstract: Modern multivariate time series forecasting primarily relies on two architectures: the Transformer with attention mechanism and Mamba. In natural language processing, an approach has been used that combines local window attention for capturing short-term dependencies and Mamba for capturing long-term dependencies, with their outputs averaged to assign equal weight to both. We find that for time-series forecasting tasks, assigning equal weight to long-term and short-term dependencies is not optimal. To mitigate this, we propose a dynamic weighting mechanism, ParallelTime Weighter, which calculates interdependent weights for long-term and short-term dependencies for each token based on the input and the model's knowledge. Furthermore, we introduce the ParallelTime architecture, which incorporates the ParallelTime Weighter mechanism to deliver state-of-the-art performance across diverse benchmarks. Our architecture demonstrates robustness, achieves lower FLOPs, requires fewer parameters, scales effectively to longer prediction horizons, and significantly outperforms existing methods. These advances highlight a promising path for future developments of parallel Attention-Mamba in time series forecasting. The implementation is readily available at: \href{https://github.com/itay1551/ParallelTime}{ParallelTime GitHub
Authors: Mathieu Godbout, Audrey Durand
Abstract: Recent work has shown that dynamic programming (DP) methods for finding static CVaR-optimal policies in Markov Decision Processes (MDPs) can fail when based on the dual formulation, yet the root cause for the failure has remained unclear. We expand on these findings by shifting focus from policy optimization to the seemingly simpler task of policy evaluation. We show that evaluating the static CVaR of a given policy can be framed as two distinct minimization problems. For their solutions to match, a set of ``risk-assignment consistency constraints'' must be satisfied, and we demonstrate that the intersection of the constraints being empty is the source of previously observed evaluation errors. Quantifying the evaluation error as the CVaR evaluation gap, we then demonstrate that the issues observed when optimizing over the dual-based CVaR DP are explained by the returned policy having a non-zero CVaR evaluation gap. We then leverage our proposed risk-assignment perspective to prove that the search for a single, uniformly optimal policy via on the dual CVaR decomposition is fundamentally limited, identifying an MDP where no single policy can be optimal across all initial risk levels.
Authors: Xu Zhang, Zhenyuan Yuan, Minghui Zhu
Abstract: In this paper, we study Byzantine-resilient federated online learning for Gaussian process regression (GPR). We develop a Byzantine-resilient federated GPR algorithm that allows a cloud and a group of agents to collaboratively learn a latent function and improve the learning performances where some agents exhibit Byzantine failures, i.e., arbitrary and potentially adversarial behavior. Each agent-based local GPR sends potentially compromised local predictions to the cloud, and the cloud-based aggregated GPR computes a global model by a Byzantine-resilient product of experts aggregation rule. Then the cloud broadcasts the current global model to all the agents. Agent-based fused GPR refines local predictions by fusing the received global model with that of the agent-based local GPR. Moreover, we quantify the learning accuracy improvements of the agent-based fused GPR over the agent-based local GPR. Experiments on a toy example and two medium-scale real-world datasets are conducted to demonstrate the performances of the proposed algorithm.
Authors: Aileen Luo, Tao Zhou, Ming Du, Martin V. Holt, Andrej Singer, Mathew J. Cherukara
Abstract: Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample's local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.
Authors: Alejandro Rodriguez-Garcia, Anindya Ghosh, Srikanth Ramaswamy
Abstract: Recent studies in continual learning have identified a transient drop in performance on mastered tasks when assimilating new ones, known as the stability gap. Such dynamics contradict the objectives of continual learning, revealing a lack of robustness in mitigating forgetting, and notably, persisting even under an ideal joint-loss regime. Examining this gap within this idealized joint training context is critical to isolate it from other sources of forgetting. We argue that it reflects an imbalance between rapid adaptation and robust retention at task boundaries, underscoring the need to investigate mechanisms that reconcile plasticity and stability within continual learning frameworks. Biological brains navigate a similar dilemma by operating concurrently on multiple timescales, leveraging neuromodulatory signals to modulate synaptic plasticity. However, artificial networks lack native multitimescale dynamics, and although optimizers like momentum-SGD and Adam introduce implicit timescale regularization, they still exhibit stability gaps. Inspired by locus coeruleus mediated noradrenergic bursts, which transiently enhance neuronal gain under uncertainty to facilitate sensory assimilation, we propose uncertainty-modulated gain dynamics - an adaptive mechanism that approximates a two-timescale optimizer and dynamically balances integration of knowledge with minimal interference on previously consolidated information. We evaluate our mechanism on domain-incremental and class-incremental variants of the MNIST and CIFAR benchmarks under joint training, demonstrating that uncertainty-modulated gain dynamics effectively attenuate the stability gap. Finally, our analysis elucidates how gain modulation replicates noradrenergic functions in cortical circuits, offering mechanistic insights into reducing stability gaps and enhance performance in continual learning tasks.
Authors: Ni Mu, Yao Luan, Qing-Shan Jia
Abstract: Multi-objective reinforcement learning (MORL) is a structured approach for optimizing tasks with multiple objectives. However, it often relies on pre-defined reward functions, which can be hard to design for balancing conflicting goals and may lead to oversimplification. Preferences can serve as more flexible and intuitive decision-making guidance, eliminating the need for complicated reward design. This paper introduces preference-based MORL (Pb-MORL), which formalizes the integration of preferences into the MORL framework. We theoretically prove that preferences can derive policies across the entire Pareto frontier. To guide policy optimization using preferences, our method constructs a multi-objective reward model that aligns with the given preferences. We further provide theoretical proof to show that optimizing this reward model is equivalent to training the Pareto optimal policy. Extensive experiments in benchmark multi-objective tasks, a multi-energy management task, and an autonomous driving task on a multi-line highway show that our method performs competitively, surpassing the oracle method, which uses the ground truth reward function. This highlights its potential for practical applications in complex real-world systems.
Authors: Xiyun Li, Yining Ding, Yuhua Jiang, Yunlong Zhao, Runpeng Xie, Shuang Xu, Yuanhua Ni, Yiqin Yang, Bo Xu
Abstract: Real-time human-artificial intelligence (AI) collaboration is crucial yet challenging, especially when AI agents must adapt to diverse and unseen human behaviors in dynamic scenarios. Existing large language model (LLM) agents often fail to accurately model the complex human mental characteristics such as domain intentions, especially in the absence of direct communication. To address this limitation, we propose a novel dual process multi-scale theory of mind (DPMT) framework, drawing inspiration from cognitive science dual process theory. Our DPMT framework incorporates a multi-scale theory of mind (ToM) module to facilitate robust human partner modeling through mental characteristic reasoning. Experimental results demonstrate that DPMT significantly enhances human-AI collaboration, and ablation studies further validate the contributions of our multi-scale ToM in the slow system.
Authors: Pankaj Yadav, Vivek Vijay
Abstract: Kolmogorov Arnold Networks (KANs) are recent architectural advancement in neural computation that offer a mathematically grounded alternative to standard neural networks. This study presents an empirical evaluation of KANs in context of class imbalanced classification, using ten benchmark datasets. We observe that KANs can inherently perform well on raw imbalanced data more effectively than Multi-Layer Perceptrons (MLPs) without any resampling strategy. However, conventional imbalance strategies fundamentally conflict with KANs mathematical structure as resampling and focal loss implementations significantly degrade KANs performance, while marginally benefiting MLPs. Crucially, KANs suffer from prohibitive computational costs without proportional performance gains. Statistical validation confirms that MLPs with imbalance techniques achieve equivalence with KANs (|d| < 0.08 across metrics) at minimal resource costs. These findings reveal that KANs represent a specialized solution for raw imbalanced data where resources permit. But their severe performance-resource tradeoffs and incompatibility with standard resampling techniques currently limits practical deployment. We identify critical research priorities as developing KAN specific architectural modifications for imbalance learning, optimizing computational efficiency, and theoretical reconciling their conflict with data augmentation. This work establishes foundational insights for next generation KAN architectures in imbalanced classification scenarios.
Authors: Jianhong Chen, Meng Zhao, Mostafa Reisi Gahrooei, Xubo Yue
Abstract: Temporal causal representation learning is a powerful tool for uncovering complex patterns in observational studies, which are often represented as low-dimensional time series. However, in many real-world applications, data are high-dimensional with varying input lengths and naturally take the form of irregular tensors. To analyze such data, irregular tensor decomposition is critical for extracting meaningful clusters that capture essential information. In this paper, we focus on modeling causal representation learning based on the transformed information. First, we present a novel causal formulation for a set of latent clusters. We then propose CaRTeD, a joint learning framework that integrates temporal causal representation learning with irregular tensor decomposition. Notably, our framework provides a blueprint for downstream tasks using the learned tensor factors, such as modeling latent structures and extracting causal information, and offers a more flexible regularization design to enhance tensor decomposition. Theoretically, we show that our algorithm converges to a stationary point. More importantly, our results fill the gap in theoretical guarantees for the convergence of state-of-the-art irregular tensor decomposition. Experimental results on synthetic and real-world electronic health record (EHR) datasets (MIMIC-III), with extensive benchmarks from both phenotyping and network recovery perspectives, demonstrate that our proposed method outperforms state-of-the-art techniques and enhances the explainability of causal representations.
Authors: Ievgenii Afanasiev, Leonid Berlyand, Mariia Kiyashko
Abstract: The paper is concerned with deformed Wigner random matrices. These matrices are closely connected with Deep Neural Networks (DNNs): weight matrices of trained DNNs could be represented in the form $R + S$, where $R$ is random and $S$ is highly correlated. The spectrum of such matrices plays a key role in rigorous underpinning of the novel pruning technique based on Random Matrix Theory. Mathematics has been done only for finite-rank matrix $S$. However, in practice rank may grow. In this paper we develop asymptotic analysis for the case of growing rank.
Authors: Riadul Islam, Dhandeep Challagundla
Abstract: Leveraging artificial intelligence (AI)-driven electronic design and automation (EDA) tools, high-performance computing, and parallelized algorithms are essential for next-generation microprocessor innovation, ensuring continued progress in computing, AI, and semiconductor technology. Machine learning-based design rule checking (DRC) and lithography hotspot detection can improve first-pass silicon success. However, conventional ML and neural network (NN)-based models use supervised learning and require a large balanced dataset (in terms of positive and negative classes) and training time. This research addresses those key challenges by proposing the first-ever unsupervised DRC violation prediction methodology. The proposed model can be built using any unbalanced dataset using only one class and set a threshold for it, then fitting any new data querying if they are within the boundary of the model for classification. This research verified the proposed model by implementing different computational cores using CMOS 28 nm technology and Synopsys Design Compiler and IC Compiler II tools. Then, layouts were divided into virtual grids to collect about 60k data for analysis and verification. The proposed method has 99.95% prediction test accuracy, while the existing support vector machine (SVM) and neural network (NN) models have 85.44\% and 98.74\% accuracy, respectively. In addition, the proposed methodology has about 26.3x and up to 6003x lower training times compared to SVM and NN-models, respectively.
Authors: Paul E. Calzada, Zahin Ibnat, Tanvir Rahman, Kamal Kandula, Danyu Lu, Sujan Kumar Saha, Farimah Farahmandi, Mark Tehranipoor
Abstract: Large Language Models (LLMs) are gaining popularity for hardware design automation, particularly through Register Transfer Level (RTL) code generation. In this work, we examine the current literature on RTL generation using LLMs and identify key requirements for training and fine-tuning datasets. We construct a robust Verilog dataset through an automated three-pronged process involving database (DB) creation and management with PostgreSQL, data collection from code hosting sites like OpenCores and GitHub, and data preprocessing to verify the codes' syntax, run logic synthesis, and extract relevant module metadata. We implement a scalable and efficient DB infrastructure to support analysis and detail our preprocessing pipeline to enforce high-quality data before DB insertion. The resulting dataset comprises 20,392 Verilog samples, 751 MB of Verilog code data, which is the largest high-quality Verilog dataset for LLM fine-tuning to our knowledge. We further evaluate the dataset, address associated challenges, and explore potential applications for future research and development in LLM-based hardware generation.
Authors: Dong Xiao, Zahra Sharif-Khodaei, M. H. Aliabadi
Abstract: Physics-guided approaches offer a promising path toward accurate and generalisable impact identification in composite structures, especially when experimental data are sparse. This paper presents a hybrid framework for impact localisation and force estimation in composite plates, combining a data-driven implementation of First-Order Shear Deformation Theory (FSDT) with machine learning and uncertainty quantification. The structural configuration and material properties are inferred from dispersion relations, while boundary conditions are identified via modal characteristics to construct a low-fidelity but physically consistent FSDT model. This model enables physics-informed data augmentation for extrapolative localisation using supervised learning. Simultaneously, an adaptive regularisation scheme derived from the same model improves the robustness of impact force reconstruction. The framework also accounts for uncertainty by propagating localisation uncertainty through the force estimation process, producing probabilistic outputs. Validation on composite plate experiments confirms the framework's accuracy, robustness, and efficiency in reducing dependence on large training datasets. The proposed method offers a scalable and transferable solution for impact monitoring and structural health management in composite aerostructures.
Authors: Rafiq Kamel, Filippo Guerranti, Simon Geisler, Stephan G\"unnemann
Abstract: Large Language Models (LLMs) are increasingly applied to tasks involving structured inputs such as graphs. Abstract Meaning Representations (AMRs), which encode rich semantics as directed graphs, offer a rigorous testbed for evaluating LLMs on text generation from such structures. Yet, current methods often arbitrarily linearize AMRs, discarding key structural cues, or rely on architectures incompatible with standard LLMs. We introduce SAFT, a structure-aware fine-tuning approach that injects graph topology into pretrained LLMs without architectural changes. We compute direction-sensitive positional encodings from the magnetic Laplacian of transformed AMRs and project them into the embedding space of the LLM. While possibly applicable to any graph-structured inputs, we focus on AMR-to-text generation as a representative and challenging benchmark. SAFT sets a new state-of-the-art on AMR 3.0 with a 3.5 BLEU improvement over baselines. Gains scale with graph complexity, highlighting the value of structure-aware representations in enhancing LLM performance. SAFT offers a general and effective pathway for bridging structured data and language models.
Authors: Chandrashekar Muniyappa, Sirisha Velampalli
Abstract: In today\'s digital world, fake news is spreading with immense speed. Its a significant concern to address. In this work, we addressed that challenge using novel graph based approach. We took dataset from Kaggle that contains real and fake news articles. To test our approach we incorporated recent covid-19 related news articles that contains both genuine and fake news that are relevant to this problem. This further enhances the dataset as well instead of relying completely on the original dataset. We propose a contextual graph-based approach to detect fake news articles. We need to convert news articles into appropriate schema, so we leverage Natural Language Processing (NLP) techniques to transform news articles into contextual graph structures. We then apply the Minimum Description Length (MDL)-based Graph-Based Anomaly Detection (GBAD) algorithm for graph mining. Graph-based methods are particularly effective for handling rich contextual data, as they enable the discovery of complex patterns that traditional query-based or statistical techniques might overlook. Our proposed approach identifies normative patterns within the dataset and subsequently uncovers anomalous patterns that deviate from these established norms.
Authors: Osama Hardan, Omar Elshenhabi, Tamer Khattab, Mohamed Mabrok
Abstract: Vision Mamba models promise transformer-level performance at linear computational cost, but their reliance on serializing 2D images into 1D sequences introduces a critical, yet overlooked, design choice: the patch scan order. In medical imaging, where modalities like brain MRI contain strong anatomical priors, this choice is non-trivial. This paper presents the first systematic study of how scan order impacts MRI segmentation. We introduce Multi-Scan 2D (MS2D), a parameter-free module for Mamba-based architectures that facilitates exploring diverse scan paths without additional computational cost. We conduct a large-scale benchmark of 21 scan strategies on three public datasets (BraTS 2020, ISLES 2022, LGG), covering over 70,000 slices. Our analysis shows conclusively that scan order is a statistically significant factor (Friedman test: $\chi^{2}_{20}=43.9, p=0.0016$), with performance varying by as much as 27 Dice points. Spatially contiguous paths -- simple horizontal and vertical rasters -- consistently outperform disjointed diagonal scans. We conclude that scan order is a powerful, cost-free hyperparameter, and provide an evidence-based shortlist of optimal paths to maximize the performance of Mamba models in medical imaging.
Authors: Arjun Rao, Esther Rolf
Abstract: A large variety of geospatial data layers is available around the world ranging from remotely-sensed raster data like satellite imagery, digital elevation models, predicted land cover maps, and human-annotated data, to data derived from environmental sensors such as air temperature or wind speed data. A large majority of machine learning models trained on satellite imagery (SatML), however, are designed primarily for optical input modalities such as multi-spectral satellite imagery. To better understand the value of using other input modalities alongside optical imagery in supervised learning settings, we generate augmented versions of SatML benchmark tasks by appending additional geographic data layers to datasets spanning classification, regression, and segmentation. Using these augmented datasets, we find that fusing additional geographic inputs with optical imagery can significantly improve SatML model performance. Benefits are largest in settings where labeled data are limited and in geographic out-of-sample settings, suggesting that multi-modal inputs may be especially valuable for data-efficiency and out-of-sample performance of SatML models. Surprisingly, we find that hard-coded fusion strategies outperform learned variants, with interesting implications for future work.
Authors: Yang Zhang, Er Jin, Yanfei Dong, Yixuan Wu, Philip Torr, Ashkan Khakzar, Johannes Stegmaier, Kenji Kawaguchi
Abstract: Recent advances in generative models have demonstrated remarkable capabilities in producing high-quality images, but their reliance on large-scale unlabeled data has raised significant safety and copyright concerns. Efforts to address these issues by erasing unwanted concepts have shown promise. However, many existing erasure methods involve excessive modifications that compromise the overall utility of the model. In this work, we address these issues by formulating a novel minimalist concept erasure objective based \emph{only} on the distributional distance of final generation outputs. Building on our formulation, we derive a tractable loss for differentiable optimization that leverages backpropagation through all generation steps in an end-to-end manner. We also conduct extensive analysis to show theoretical connections with other models and methods. To improve the robustness of the erasure, we incorporate neuron masking as an alternative to model fine-tuning. Empirical evaluations on state-of-the-art flow-matching models demonstrate that our method robustly erases concepts without degrading overall model performance, paving the way for safer and more responsible generative models.
Authors: Kundeshwar Pundalik, Piyush Sawarkar, Nihar Sahoo, Abhishek Shinde, Prateek Chanda, Vedant Goswami, Ajay Nagpal, Atul Singh, Viraj Thakur, Vijay Dewane, Aamod Thakur, Bhargav Patel, Smita Gautam, Bhagwan Panditi, Shyam Pawar, Madhav Kotcha, Suraj Racha, Saral Sureka, Pankaj Singh, Rishi Bal, Rohit Saluja, Ganesh Ramakrishnan
Abstract: Large Language Models (LLMs) have emerged as powerful general-purpose reasoning systems, yet their development remains dominated by English-centric data, architectures, and optimization paradigms. This exclusionary design results in structural under-representation of linguistically diverse regions such as India, where over 20 official languages and 100+ dialects coexist alongside phenomena like code-switching and diglossia. We introduce PARAM-1, a 2.9B parameter decoder-only, text-only language model trained from scratch with an explicit architectural and linguistic focus on Indian diversity. PARAM-1 is trained on a bilingual dataset consisting of only Hindi and English, constructed with a strong focus on fact-rich, high-quality content. It is guided by three core principles: equitable representation of Indic languages through a 25% corpus allocation; tokenization fairness via a SentencePiece tokenizer adapted to Indian morphological structures; and culturally aligned evaluation benchmarks across IndicQA, code-mixed reasoning, and socio-linguistic robustness tasks. By embedding diversity at the pretraining level-rather than deferring it to post-hoc alignment-PARAM-1 offers a design-first blueprint for equitable foundation modeling. Our results demonstrate that it serves as both a competent general-purpose model and a robust baseline for India-centric applications.
Authors: Shreya Kadambi, Risheek Garrepalli, Shubhankar Borse, Munawar Hyatt, Fatih Porikli
Abstract: Despite the remarkable success of diffusion models in text-to-image generation, their effectiveness in grounded visual editing and compositional control remains challenging. Motivated by advances in self-supervised learning and in-context generative modeling, we propose a series of simple yet powerful design choices that significantly enhance diffusion model capacity for structured, controllable generation and editing. We introduce Masking-Augmented Diffusion with Inference-Time Scaling (MADI), a framework that improves the editability, compositionality and controllability of diffusion models through two core innovations. First, we introduce Masking-Augmented gaussian Diffusion (MAgD), a novel training strategy with dual corruption process which combines standard denoising score matching and masked reconstruction by masking noisy input from forward process. MAgD encourages the model to learn discriminative and compositional visual representations, thus enabling localized and structure-aware editing. Second, we introduce an inference-time capacity scaling mechanism based on Pause Tokens, which act as special placeholders inserted into the prompt for increasing computational capacity at inference time. Our findings show that adopting expressive and dense prompts during training further enhances performance, particularly for MAgD. Together, these contributions in MADI substantially enhance the editability of diffusion models, paving the way toward their integration into more general-purpose, in-context generative diffusion architectures.
Authors: Morteza Bodaghi, Majid Hosseini, Raju Gottumukkala, Ravi Teja Bhupatiraju, Iftikhar Ahmad, Moncef Gabbouj
Abstract: In this study, we present a comprehensive public dataset for driver drowsiness detection, integrating multimodal signals of facial, behavioral, and biometric indicators. Our dataset includes 3D facial video using a depth camera, IR camera footage, posterior videos, and biometric signals such as heart rate, electrodermal activity, blood oxygen saturation, skin temperature, and accelerometer data. This data set provides grip sensor data from the steering wheel and telemetry data from the American truck simulator game to provide more information about drivers' behavior while they are alert and drowsy. Drowsiness levels were self-reported every four minutes using the Karolinska Sleepiness Scale (KSS). The simulation environment consists of three monitor setups, and the driving condition is completely like a car. Data were collected from 19 subjects (15 M, 4 F) in two conditions: when they were fully alert and when they exhibited signs of sleepiness. Unlike other datasets, our multimodal dataset has a continuous duration of 40 minutes for each data collection session per subject, contributing to a total length of 1,400 minutes, and we recorded gradual changes in the driver state rather than discrete alert/drowsy labels. This study aims to create a comprehensive multimodal dataset of driver drowsiness that captures a wider range of physiological, behavioral, and driving-related signals. The dataset will be available upon request to the corresponding author.
Authors: Ishant Chintapatla, Kazuma Choji, Naaisha Agarwal, Andrew Lin, Hannah You, Charles Duong, Kevin Zhu, Sean O'Brien, Vasu Sharma
Abstract: Recently, many benchmarks and datasets have been developed to evaluate Vision-Language Models (VLMs) using visual question answering (VQA) pairs, and models have shown significant accuracy improvements. However, these benchmarks rarely test the model's ability to accurately complete visual entailment, for instance, accepting or refuting a hypothesis based on the image. To address this, we propose COREVQA (Crowd Observations and Reasoning Entailment), a benchmark of 5608 image and synthetically generated true/false statement pairs, with images derived from the CrowdHuman dataset, to provoke visual entailment reasoning on challenging crowded images. Our results show that even the top-performing VLMs achieve accuracy below 80%, with other models performing substantially worse (39.98%-69.95%). This significant performance gap reveals key limitations in VLMs' ability to reason over certain types of image-question pairs in crowded scenes.
Authors: Alessandro Pistola, Valentina Orru', Nicolo' Marchetti, Marco Roccetti
Abstract: By upgrading an existing deep learning model with the knowledge provided by one of the oldest sets of grayscale satellite imagery, known as CORONA, we improved the AI model attitude towards the automatic identification of archaeological sites in an environment which has been completely transformed in the last five decades, including the complete destruction of many of those same sites. The initial Bing based convolutional network model was retrained using CORONA satellite imagery for the district of Abu Ghraib, west of Baghdad, central Mesopotamian floodplain. The results were twofold and surprising. First, the detection precision obtained on the area of interest increased sensibly: in particular, the Intersection over Union (IoU) values, at the image segmentation level, surpassed 85 percent, while the general accuracy in detecting archeological sites reached 90 percent. Second, our retrained model allowed the identification of four new sites of archaeological interest (confirmed through field verification), previously not identified by archaeologists with traditional techniques. This has confirmed the efficacy of using AI techniques and the CORONA imagery from the 1960 to discover archaeological sites currently no longer visible, a concrete breakthrough with significant consequences for the study of landscapes with vanishing archaeological evidence induced by anthropization
Authors: Malte Hoffmann
Abstract: Deep learning has revolutionized neuroimage analysis by delivering unprecedented speed and accuracy. However, the narrow scope of many training datasets constrains model robustness and generalizability. This challenge is particularly acute in magnetic resonance imaging (MRI), where image appearance varies widely across pulse sequences and scanner hardware. A recent domain-randomization strategy addresses the generalization problem by training deep neural networks on synthetic images with randomized intensities and anatomical content. By generating diverse data from anatomical segmentation maps, the approach enables models to accurately process image types unseen during training, without retraining or fine-tuning. It has demonstrated effectiveness across modalities including MRI, computed tomography, positron emission tomography, and optical coherence tomography, as well as beyond neuroimaging in ultrasound, electron and fluorescence microscopy, and X-ray microtomography. This tutorial paper reviews the principles, implementation, and potential of the synthesis-driven training paradigm. It highlights key benefits, such as improved generalization and resistance to overfitting, while discussing trade-offs such as increased computational demands. Finally, the article explores practical considerations for adopting the technique, aiming to accelerate the development of generalizable tools that make deep learning more accessible to domain experts without extensive computational resources or machine learning knowledge.
Authors: Vijay K. Dubey (The University of Texas at Austin), Collin E. Haese (The University of Texas at Austin), Osman G\"ultekin (The University of Texas at Austin), David Dalton (University of Glasgow), Manuel K. Rausch (The University of Texas at Austin), Jan N. Fuhg (The University of Texas at Austin)
Abstract: Surrogate models for the rapid inference of nonlinear boundary value problems in mechanics are helpful in a broad range of engineering applications. However, effective surrogate modeling of applications involving the contact of deformable bodies, especially in the context of varying geometries, is still an open issue. In particular, existing methods are confined to rigid body contact or, at best, contact between rigid and soft objects with well-defined contact planes. Furthermore, they employ contact or collision detection filters that serve as a rapid test but use only the necessary and not sufficient conditions for detection. In this work, we present a graph neural network architecture that utilizes continuous collision detection and, for the first time, incorporates sufficient conditions designed for contact between soft deformable bodies. We test its performance on two benchmarks, including a problem in soft tissue mechanics of predicting the closed state of a bioprosthetic aortic valve. We find a regularizing effect on adding additional contact terms to the loss function, leading to better generalization of the network. These benefits hold for simple contact at similar planes and element normal angles, and complex contact at differing planes and element normal angles. We also demonstrate that the framework can handle varying reference geometries. However, such benefits come with high computational costs during training, resulting in a trade-off that may not always be favorable. We quantify the training cost and the resulting inference speedups on various hardware architectures. Importantly, our graph neural network implementation results in up to a thousand-fold speedup for our benchmark problems at inference.
Authors: Sara Avesani, Gianluca Giacchi, Michael Multerer
Abstract: Inspired by edge detection based on the decay behavior of wavelet coefficients, we introduce a (near) linear-time algorithm for detecting the local regularity in non-uniformly sampled multivariate signals. Our approach quantifies regularity within the framework of microlocal spaces introduced by Jaffard. The central tool in our analysis is the fast samplet transform, a distributional wavelet transform tailored to scattered data. We establish a connection between the decay of samplet coefficients and the pointwise regularity of multivariate signals. As a by product, we derive decay estimates for functions belonging to classical H\"older spaces and Sobolev-Slobodeckij spaces. While traditional wavelets are effective for regularity detection in low-dimensional structured data, samplets demonstrate robust performance even for higher dimensional and scattered data. To illustrate our theoretical findings, we present extensive numerical studies detecting local regularity of one-, two- and three-dimensional signals, ranging from non-uniformly sampled time series over image segmentation to edge detection in point clouds.
Authors: Imane Hamzaoui, Riyadh Baghdadi
Abstract: Bio-inspired neural networks are attractive for their adversarial robustness, energy frugality, and closer alignment with cortical physiology, yet they often lag behind back-propagation (BP) based models in accuracy and ability to scale. We show that allowing the use of different bio-inspired learning rules in different layers, discovered automatically by a tailored neural-architecture-search (NAS) procedure, bridges this gap. Starting from standard NAS baselines, we enlarge the search space to include bio-inspired learning rules and use NAS to find the best architecture and learning rule to use in each layer. We show that neural networks that use different bio-inspired learning rules for different layers have better accuracy than those that use a single rule across all the layers. The resulting NN that uses a mix of bio-inspired learning rules sets new records for bio-inspired models: 95.16% on CIFAR-10, 76.48% on CIFAR-100, 43.42% on ImageNet16-120, and 60.51% top-1 on ImageNet. In some regimes, they even surpass comparable BP-based networks while retaining their robustness advantages. Our results suggest that layer-wise diversity in learning rules allows better scalability and accuracy, and motivates further research on mixing multiple bio-inspired learning rules in the same network.
Authors: Steven Lamp, Jason D. Hiser, Anh Nguyen-Tuong, Jack W. Davidson
Abstract: Cybersecurity simulation environments, such as cyber ranges, honeypots, and sandboxes, require realistic human behavior to be effective, yet no quantitative method exists to assess the behavioral fidelity of synthetic user personas. This paper presents PHASE (Passive Human Activity Simulation Evaluation), a machine learning framework that analyzes Zeek connection logs and distinguishes human from non-human activity with over 90\% accuracy. PHASE operates entirely passively, relying on standard network monitoring without any user-side instrumentation or visible signs of surveillance. All network activity used for machine learning is collected via a Zeek network appliance to avoid introducing unnecessary network traffic or artifacts that could disrupt the fidelity of the simulation environment. The paper also proposes a novel labeling approach that utilizes local DNS records to classify network traffic, thereby enabling machine learning analysis. Furthermore, we apply SHAP (SHapley Additive exPlanations) analysis to uncover temporal and behavioral signatures indicative of genuine human users. In a case study, we evaluate a synthetic user persona and identify distinct non-human patterns that undermine behavioral realism. Based on these insights, we develop a revised behavioral configuration that significantly improves the human-likeness of synthetic activity yielding a more realistic and effective synthetic user persona.
Authors: Bhumika Laxman Sadbhave, Philipp Vaeth, Denise Dejon, Gunther Schorcht, Magda Gregorov\'a
Abstract: Satellite Image Time Series (SITS) data has proven effective for agricultural tasks due to its rich spectral and temporal nature. In this study, we tackle the task of stress detection in sugar-beet fields using a fully unsupervised approach. We propose a 3D convolutional autoencoder model to extract meaningful features from Sentinel-2 image sequences, combined with acquisition-date-specific temporal encodings to better capture the growth dynamics of sugar-beets. The learned representations are used in a downstream clustering task to separate stressed from healthy fields. The resulting stress detection system can be directly applied to data from different years, offering a practical and accessible tool for stress detection in sugar-beets.
Authors: Alexander Kolpakov
Abstract: We develop a framework for dualizing the Kolmogorov structure function $h_x(\alpha)$, which then allows using computable complexity proxies. We establish a mathematical analogy between information-theoretic constructs and statistical mechanics, introducing a suitable partition function and free energy functional. We explicitly prove the Legendre-Fenchel duality between the structure function and free energy, showing detailed balance of the Metropolis kernel, and interpret acceptance probabilities as information-theoretic scattering amplitudes. A susceptibility-like variance of model complexity is shown to peak precisely at loss-complexity trade-offs interpreted as phase transitions. Practical experiments with linear and tree-based regression models verify these theoretical predictions, explicitly demonstrating the interplay between the model complexity, generalization, and overfitting threshold.
Authors: David Poole
Abstract: AI seems to be taking over the world with systems that model pixels, words, and phonemes. The world is arguably made up, not of pixels, words, and phonemes but of entities (objects, things, including events) with properties and relations among them. Surely we should model these, not the perception or description of them. You might suspect that concentrating on modeling words and pixels is because all of the (valuable) data in the world is in terms of text and images. If you look into almost any company you will find their most valuable data is in spreadsheets, databases and other relational formats. These are not the form that are studied in introductory machine learning, but are full of product numbers, student numbers, transaction numbers and other identifiers that can't be interpreted naively as numbers. The field that studies this sort of data has various names including relational learning, statistical relational AI, and many others. This paper explains why relational learning is not taking over the world -- except in a few cases with restricted relations -- and what needs to be done to bring it to it's rightful prominence.
Authors: Hao Tuo, Yan Li, Xuanning Hu, Haishi Zhao, Xueyan Liu, Bo Yang
Abstract: Combinatorial optimization algorithm is essential in computer-aided drug design by progressively exploring chemical space to design lead compounds with high affinity to target protein. However current methods face inherent challenges in integrating domain knowledge, limiting their performance in identifying lead compounds with novel and valid binding mode. Here, we propose AutoLeadDesign, a lead compounds design framework that inspires extensive domain knowledge encoded in large language models with chemical fragments to progressively implement efficient exploration of vast chemical space. The comprehensive experiments indicate that AutoLeadDesign outperforms baseline methods. Significantly, empirical lead design campaigns targeting two clinically relevant targets (PRMT5 and SARS-CoV-2 PLpro) demonstrate AutoLeadDesign's competence in de novo generation of lead compounds achieving expert-competitive design efficacy. Structural analysis further confirms their mechanism-validated inhibitory patterns. By tracing the process of design, we find that AutoLeadDesign shares analogous mechanisms with fragment-based drug design which traditionally rely on the expert decision-making, further revealing why it works. Overall, AutoLeadDesign offers an efficient approach for lead compounds design, suggesting its potential utility in drug design.
Authors: Sahar Ghoflsaz Ghinani, Elaheh Sadredini
Abstract: Federated Learning (FL) enables collaborative model training without centralizing client data, making it attractive for privacy-sensitive domains. While existing approaches employ cryptographic techniques such as homomorphic encryption, differential privacy, or secure multiparty computation to mitigate inference attacks-including model inversion, membership inference, and gradient leakage-they often suffer from high computational, communication, or memory overheads. Moreover, many methods overlook the confidentiality of the global model itself, which may be proprietary and sensitive. These challenges limit the practicality of secure FL, especially in cross-silo deployments involving large datasets and strict compliance requirements. We present FuSeFL, a fully secure and scalable FL scheme designed for cross-silo settings. FuSeFL decentralizes training across client pairs using lightweight secure multiparty computation (MPC), while confining the server's role to secure aggregation. This design eliminates server bottlenecks, avoids data offloading, and preserves full confidentiality of data, model, and updates throughout training. FuSeFL defends against inference threats, achieves up to 95% lower communication latency and 50% lower server memory usage, and improves accuracy over prior secure FL solutions, demonstrating strong security and efficiency at scale.
Authors: Amro Abdalla, Ismail Shaheen, Dan DeGenaro, Rupayan Mallick, Bogdan Raita, Sarah Adel Bargal
Abstract: We present GIFT: a {G}radient-aware {I}mmunization technique to defend diffusion models against malicious {F}ine-{T}uning while preserving their ability to generate safe content. Existing safety mechanisms like safety checkers are easily bypassed, and concept erasure methods fail under adversarial fine-tuning. GIFT addresses this by framing immunization as a bi-level optimization problem: the upper-level objective degrades the model's ability to represent harmful concepts using representation noising and maximization, while the lower-level objective preserves performance on safe data. GIFT achieves robust resistance to malicious fine-tuning while maintaining safe generative quality. Experimental results show that our method significantly impairs the model's ability to re-learn harmful concepts while maintaining performance on safe content, offering a promising direction for creating inherently safer generative models resistant to adversarial fine-tuning attacks.
Authors: Shivakanth Sujit, Luca Nunziante, Dan Ogawa Lillrank, Rousslan Fernand Julien Dossa, Kai Arulkumaran
Abstract: In this work we extend the low-cost GELLO teleoperation system, initially designed for joint position control, with additional force information. Our first extension is to implement force feedback, allowing users to feel resistance when interacting with the environment. Our second extension is to add force information into the data collection process and training of imitation learning models. We validate our additions by implementing these on a GELLO system with a Franka Panda arm as the follower robot, performing a user study, and comparing the performance of policies trained with and without force information on a range of simulated and real dexterous manipulation tasks. Qualitatively, users with robotics experience preferred our controller, and the addition of force inputs improved task success on the majority of tasks.
Authors: Niveen O. Jaffal, Mohammed Alkhanafseh, David Mohaisen
Abstract: Large Language Models (LLMs) are transforming cybersecurity by enabling intelligent, adaptive, and automated approaches to threat detection, vulnerability assessment, and incident response. With their advanced language understanding and contextual reasoning, LLMs surpass traditional methods in tackling challenges across domains such as IoT, blockchain, and hardware security. This survey provides a comprehensive overview of LLM applications in cybersecurity, focusing on two core areas: (1) the integration of LLMs into key cybersecurity domains, and (2) the vulnerabilities of LLMs themselves, along with mitigation strategies. By synthesizing recent advancements and identifying key limitations, this work offers practical insights and strategic recommendations for leveraging LLMs to build secure, scalable, and future-ready cyber defense systems.
Authors: Sen Lu, Xiaoyu Zhang, Mingtao Hu, Eric Yeu-Jer Lee, Soohyeon Kim, Wei D. Lu
Abstract: A grand challenge in modern neuroscience is to bridge the gap between the detailed mapping of microscale neural circuits and a mechanistic understanding of cognitive functions. While extensive knowledge exists about neuronal connectivity and biophysics, a significant gap remains in how these elements combine to produce flexible, learned behaviors. Here, we propose that a framework based on State-Space Models (SSMs), an emerging class of deep learning architectures, can bridge this gap. We argue that the differential equations governing elements in an SSM are conceptually consistent with the biophysical dynamics of neurons, while the combined dynamics in the model lead to emergent behaviors observed in experimental neuroscience. We test this framework by training an S5 model--a specific SSM variant employing a diagonal state transition matrix--on temporal discrimination tasks with reinforcement learning (RL). We demonstrate that the model spontaneously develops neural representations that strikingly mimic biological 'time cells'. We reveal that these cells emerge from a simple generative principle: learned rotational dynamics of hidden state vectors in the complex plane. This single mechanism unifies the emergence of time cells, ramping activity, and oscillations/traveling waves observed in numerous experiments. Furthermore, we show that this rotational dynamics generalizes beyond interval discriminative tasks to abstract event-counting tasks that were considered foundational for performing complex cognitive tasks. Our findings position SSMs as a compelling framework that connects single-neuron dynamics to cognitive phenomena, offering a unifying and computationally tractable theoretical ground for temporal learning in the brain.
Authors: Nikola Pavlovic, Sudeep Salgia, Qing Zhao
Abstract: We consider the problem of contextual kernel bandits with stochastic contexts, where the underlying reward function belongs to a known Reproducing Kernel Hilbert Space. We study this problem under an additional constraint of Differential Privacy, where the agent needs to ensure that the sequence of query points is differentially private with respect to both the sequence of contexts and rewards. We propose a novel algorithm that achieves the state-of-the-art cumulative regret of $\widetilde{\mathcal{O}}(\sqrt{\gamma_TT}+\frac{\gamma_T}{\varepsilon_{\mathrm{DP}}})$ and $\widetilde{\mathcal{O}}(\sqrt{\gamma_TT}+\frac{\gamma_T\sqrt{T}}{\varepsilon_{\mathrm{DP}}})$ over a time horizon of $T$ in the joint and local models of differential privacy, respectively, where $\gamma_T$ is the effective dimension of the kernel and $\varepsilon_{\mathrm{DP}} > 0$ is the privacy parameter. The key ingredient of the proposed algorithm is a novel private kernel-ridge regression estimator which is based on a combination of private covariance estimation and private random projections. It offers a significantly reduced sensitivity compared to its classical counterpart while maintaining a high prediction accuracy, allowing our algorithm to achieve the state-of-the-art performance guarantees.
Authors: Xiao Wang, Qian Zhu, Shujuan Wu, Bo Jiang, Shiliang Zhang, Yaowei Wang, Yonghong Tian, Bin Luo
Abstract: Recent researchers have proposed using event cameras for person re-identification (ReID) due to their promising performance and better balance in terms of privacy protection, event camera-based person ReID has attracted significant attention. Currently, mainstream event-based person ReID algorithms primarily focus on fusing visible light and event stream, as well as preserving privacy. Although significant progress has been made, these methods are typically trained and evaluated on small-scale or simulated event camera datasets, making it difficult to assess their real identification performance and generalization ability. To address the issue of data scarcity, this paper introduces a large-scale RGB-event based person ReID dataset, called EvReID. The dataset contains 118,988 image pairs and covers 1200 pedestrian identities, with data collected across multiple seasons, scenes, and lighting conditions. We also evaluate 15 state-of-the-art person ReID algorithms, laying a solid foundation for future research in terms of both data and benchmarking. Based on our newly constructed dataset, this paper further proposes a pedestrian attribute-guided contrastive learning framework to enhance feature learning for person re-identification, termed TriPro-ReID. This framework not only effectively explores the visual features from both RGB frames and event streams, but also fully utilizes pedestrian attributes as mid-level semantic features. Extensive experiments on the EvReID dataset and MARS datasets fully validated the effectiveness of our proposed RGB-Event person ReID framework. The benchmark dataset and source code will be released on https://github.com/Event-AHU/Neuromorphic_ReID
Authors: Chuheng Wei, Ziye Qin, Walter Zimmer, Guoyuan Wu, Matthew J. Barth
Abstract: Real-world Vehicle-to-Everything (V2X) cooperative perception systems often operate under heterogeneous sensor configurations due to cost constraints and deployment variability across vehicles and infrastructure. This heterogeneity poses significant challenges for feature fusion and perception reliability. To address these issues, we propose HeCoFuse, a unified framework designed for cooperative perception across mixed sensor setups where nodes may carry Cameras (C), LiDARs (L), or both. By introducing a hierarchical fusion mechanism that adaptively weights features through a combination of channel-wise and spatial attention, HeCoFuse can tackle critical challenges such as cross-modality feature misalignment and imbalanced representation quality. In addition, an adaptive spatial resolution adjustment module is employed to balance computational cost and fusion effectiveness. To enhance robustness across different configurations, we further implement a cooperative learning strategy that dynamically adjusts fusion type based on available modalities. Experiments on the real-world TUMTraf-V2X dataset demonstrate that HeCoFuse achieves 43.22% 3D mAP under the full sensor configuration (LC+LC), outperforming the CoopDet3D baseline by 1.17%, and reaches an even higher 43.38% 3D mAP in the L+LC scenario, while maintaining 3D mAP in the range of 21.74% to 43.38% across nine heterogeneous sensor configurations. These results, validated by our first-place finish in the CVPR 2025 DriveX challenge, establish HeCoFuse as the current state-of-the-art on TUM-Traf V2X dataset while demonstrating robust performance across diverse sensor deployments.
Authors: Emma Rapoport, Edith Cohen, Uri Stemmer
Abstract: Most work on adaptive data analysis assumes that samples in the dataset are independent. When correlations are allowed, even the non-adaptive setting can become intractable, unless some structural constraints are imposed. To address this, Bassily and Freund [2016] introduced the elegant framework of concentrated queries, which requires the analyst to restrict itself to queries that are concentrated around their expected value. While this assumption makes the problem trivial in the non-adaptive setting, in the adaptive setting it remains quite challenging. In fact, all known algorithms in this framework support significantly fewer queries than in the independent case: At most $O(n)$ queries for a sample of size $n$, compared to $O(n^2)$ in the independent setting. In this work, we prove that this utility gap is inherent under the current formulation of the concentrated queries framework, assuming some natural conditions on the algorithm. Additionally, we present a simplified version of the best-known algorithms that match our impossibility result.
Authors: Jing Chang, Chang Liu, Jinbin Huang, Rui Mao, Jianbin Qin
Abstract: Data preparation is a foundational yet notoriously challenging component of the machine learning lifecycle, characterized by a vast combinatorial search space of potential operator sequences. While reinforcement learning (RL) offers a promising direction, existing approaches are inefficient as they fail to capture the structured, hierarchical nature of the problem. We argue that Hierarchical Reinforcement Learning (HRL), a paradigm that has been successful in other domains, provides a conceptually ideal yet previously unexplored framework for this task. However, a naive HRL implementation with a `hard hierarchy' is prone to suboptimal, irreversible decisions. To address this, we introduce CogniQ-H, the first framework to implement a soft hierarchical paradigm for robust, end-to-end automated data preparation. CogniQ-H formulates action selection as a Bayesian inference problem. A high-level strategic prior, generated by a Large Language Model (LLM), guides exploration probabilistically. This prior is synergistically combined with a fine-grained operator quality score from a supervised Learning-to-Rank (LTR) model and a long-term value estimate from the agent's own Q-function. This hybrid architecture allows CogniQ-H to balance strategic guidance with adaptive, evidence-based decision-making. Through extensive experiments on 18 diverse datasets spanning multiple domains, we demonstrate that CogniQ-H achieves up to 13.9\% improvement in pipeline quality and 2.8$\times$ faster convergence compared to state-of-the-art RL-based methods.
Authors: Jing Chang, Chang Liu, Jinbin Huang, Rui Mao, Jianbin Qin
Abstract: Automated data preparation is crucial for democratizing machine learning, yet existing reinforcement learning (RL) based approaches suffer from inefficient exploration in the vast space of possible preprocessing pipelines. We present LLaPipe, a novel framework that addresses this exploration bottleneck by integrating Large Language Models (LLMs) as intelligent policy advisors. Unlike traditional methods that rely solely on statistical features and blind trial-and-error, LLaPipe leverages the semantic understanding capabilities of LLMs to provide contextually relevant exploration guidance. Our framework introduces three key innovations: (1) an LLM Policy Advisor that analyzes dataset semantics and pipeline history to suggest promising preprocessing operations, (2) an Experience Distillation mechanism that mines successful patterns from past pipelines and transfers this knowledge to guide future exploration, and (3) an Adaptive Advisor Triggering strategy (Advisor\textsuperscript{+}) that dynamically determines when LLM intervention is most beneficial, balancing exploration effectiveness with computational cost. Through extensive experiments on 18 diverse datasets spanning multiple domains, we demonstrate that LLaPipe achieves up to 22.4\% improvement in pipeline quality and 2.3$\times$ faster convergence compared to state-of-the-art RL-based methods, while maintaining computational efficiency through selective LLM usage (averaging only 19.0\% of total exploration steps).
Authors: Julia Laubmann, Johannes Reschke
Abstract: In today's digital age, concerns about the dangers of AI-generated images are increasingly common. One powerful tool in this domain is StyleGAN (style-based generative adversarial networks), a generative adversarial network capable of producing highly realistic synthetic faces. To gain a deeper understanding of how such a model operates, this work focuses on analyzing the inner workings of StyleGAN's generator component. Key architectural elements and techniques, such as the Equalized Learning Rate, are explored in detail to shed light on the model's behavior. A StyleGAN model is trained using the PyTorch framework, enabling direct inspection of its learned weights. Through pruning, it is revealed that a significant number of these weights can be removed without drastically affecting the output, leading to reduced computational requirements. Moreover, the role of the latent vector -- which heavily influences the appearance of the generated faces -- is closely examined. Global alterations to this vector primarily affect aspects like color tones, while targeted changes to individual dimensions allow for precise manipulation of specific facial features. This ability to finetune visual traits is not only of academic interest but also highlights a serious ethical concern: the potential misuse of such technology. Malicious actors could exploit this capability to fabricate convincing fake identities, posing significant risks in the context of digital deception and cybercrime.
Authors: Guillaume Zambrano
Abstract: This study examines the role of human judges in legal decision-making by using machine learning to predict child physical custody outcomes in French appellate courts. Building on the legal realism-formalism debate, we test whether individual judges' decision-making patterns significantly influence case outcomes, challenging the assumption that judges are neutral variables that apply the law uniformly. To ensure compliance with French privacy laws, we implement a strict pseudonymization process. Our analysis uses 18,937 living arrangements rulings extracted from 10,306 cases. We compare models trained on individual judges' past rulings (specialist models) with a judge-agnostic model trained on aggregated data (generalist models). The prediction pipeline is a hybrid approach combining large language models (LLMs) for structured feature extraction and ML models for outcome prediction (RF, XGB and SVC). Our results show that specialist models consistently achieve higher predictive accuracy than the general model, with top-performing models reaching F1 scores as high as 92.85%, compared to the generalist model's 82.63% trained on 20x to 100x more samples. Specialist models capture stable individual patterns that are not transferable to other judges. In-Domain and Cross-Domain validity tests provide empirical support for legal realism, demonstrating that judicial identity plays a measurable role in legal outcomes. All data and code used will be made available.
Authors: Abhijit Sen, Giridas Maiti, Bikram K. Parida, Bhanu P. Mishra, Mahima Arya, Denys I. Bondar
Abstract: Feature engineering continues to play a critical role in image classification, particularly when interpretability and computational efficiency are prioritized over deep learning models with millions of parameters. In this study, we revisit classical machine learning based image classification through a novel approach centered on Permutation Entropy (PE), a robust and computationally lightweight measure traditionally used in time series analysis but rarely applied to image data. We extend PE to two-dimensional images and propose a multiscale, multi-orientation entropy-based feature extraction approach that characterizes spatial order and complexity along rows, columns, diagonals, anti-diagonals, and local patches of the image. To enhance the discriminatory power of the entropy features, we integrate two classic image descriptors: the Histogram of Oriented Gradients (HOG) to capture shape and edge structure, and Local Binary Patterns (LBP) to encode micro-texture of an image. The resulting hand-crafted feature set, comprising of 780 dimensions, is used to train Support Vector Machine (SVM) classifiers optimized through grid search. The proposed approach is evaluated on multiple benchmark datasets, including Fashion-MNIST, KMNIST, EMNIST, and CIFAR-10, where it delivers competitive classification performance without relying on deep architectures. Our results demonstrate that the fusion of PE with HOG and LBP provides a compact, interpretable, and effective alternative to computationally expensive and limited interpretable deep learning models. This shows a potential of entropy-based descriptors in image classification and contributes a lightweight and generalizable solution to interpretable machine learning in image classification and computer vision.
Authors: Hosein Azarbonyad, Zi Long Zhu, Georgios Cheirmpos, Zubair Afzal, Vikrant Yadav, Georgios Tsatsaronis
Abstract: When deciding to read an article or incorporate it into their research, scholars often seek to quickly identify and understand its main ideas. In this paper, we aim to extract these key concepts and contributions from scientific articles in the form of Question and Answer (QA) pairs. We propose two distinct approaches for generating QAs. The first approach involves selecting salient paragraphs, using a Large Language Model (LLM) to generate questions, ranking these questions by the likelihood of obtaining meaningful answers, and subsequently generating answers. This method relies exclusively on the content of the articles. However, assessing an article's novelty typically requires comparison with the existing literature. Therefore, our second approach leverages a Knowledge Graph (KG) for QA generation. We construct a KG by fine-tuning an Entity Relationship (ER) extraction model on scientific articles and using it to build the graph. We then employ a salient triplet extraction method to select the most pertinent ERs per article, utilizing metrics such as the centrality of entities based on a triplet TF-IDF-like measure. This measure assesses the saliency of a triplet based on its importance within the article compared to its prevalence in the literature. For evaluation, we generate QAs using both approaches and have them assessed by Subject Matter Experts (SMEs) through a set of predefined metrics to evaluate the quality of both questions and answers. Our evaluations demonstrate that the KG-based approach effectively captures the main ideas discussed in the articles. Furthermore, our findings indicate that fine-tuning the ER extraction model on our scientific corpus is crucial for extracting high-quality triplets from such documents.
Authors: Martin V. Vejling, Shashi Raj Pandey, Christophe A. N. Biscio, Petar Popovski
Abstract: The amount of quality data in many machine learning tasks is limited to what is available locally to data owners. The set of quality data can be expanded through trading or sharing with external data agents. However, data buyers need quality guarantees before purchasing, as external data may be contaminated or irrelevant to their specific learning task. Previous works primarily rely on distributional assumptions about data from different agents, relegating quality checks to post-hoc steps involving costly data valuation procedures. We propose a distribution-free, contamination-aware data-sharing framework that identifies external data agents whose data is most valuable for model personalization. To achieve this, we introduce novel two-sample testing procedures, grounded in rigorous theoretical foundations for conformal outlier detection, to determine whether an agent's data exceeds a contamination threshold. The proposed tests, termed conformal data contamination tests, remain valid under arbitrary contamination levels while enabling false discovery rate control via the Benjamini-Hochberg procedure. Empirical evaluations across diverse collaborative learning scenarios demonstrate the robustness and effectiveness of our approach. Overall, the conformal data contamination test distinguishes itself as a generic procedure for aggregating data with statistically rigorous quality guarantees.
Authors: Mehul Anand, Shishir Kolathaya
Abstract: Synthesising safe controllers from visual data typically requires extensive supervised labelling of safety-critical data, which is often impractical in real-world settings. Recent advances in world models enable reliable prediction in latent spaces, opening new avenues for scalable and data-efficient safe control. In this work, we introduce a semi-supervised framework that leverages control barrier certificates (CBCs) learned in the latent space of a world model to synthesise safe visuomotor policies. Our approach jointly learns a neural barrier function and a safe controller using limited labelled data, while exploiting the predictive power of modern vision transformers for latent dynamics modelling.
Authors: James A. D. Binnie, Pawe{\l} D{\l}otko, John Harvey, Jakub Malinowski, Ka Man Yim
Abstract: It is a standard assumption that datasets in high dimension have an internal structure which means that they in fact lie on, or near, subsets of a lower dimension. In many instances it is important to understand the real dimension of the data, hence the complexity of the dataset at hand. A great variety of dimension estimators have been developed to find the intrinsic dimension of the data but there is little guidance on how to reliably use these estimators. This survey reviews a wide range of dimension estimation methods, categorising them by the geometric information they exploit: tangential estimators which detect a local affine structure; parametric estimators which rely on dimension-dependent probability distributions; and estimators which use topological or metric invariants. The paper evaluates the performance of these methods, as well as investigating varying responses to curvature and noise. Key issues addressed include robustness to hyperparameter selection, sample size requirements, accuracy in high dimensions, precision, and performance on non-linear geometries. In identifying the best hyperparameters for benchmark datasets, overfitting is frequent, indicating that many estimators may not generalise well beyond the datasets on which they have been tested.
Authors: Jacob C Walker, Pedro V\'elez, Luisa Polania Cabrera, Guangyao Zhou, Rishabh Kabra, Carl Doersch, Maks Ovsjanikov, Jo\~ao Carreira, Shiry Ginosar
Abstract: Forecasting what will happen next is a critical skill for general-purpose systems that plan or act in the world at different levels of abstraction. In this paper, we identify a strong correlation between a vision model's perceptual ability and its generalist forecasting performance over short time horizons. This trend holds across a diverse set of pretrained models-including those trained generatively-and across multiple levels of abstraction, from raw pixels to depth, point tracks, and object motion. The result is made possible by a novel generalist forecasting framework that operates on any frozen vision backbone: we train latent diffusion models to forecast future features in the frozen representation space, which are then decoded via lightweight, task-specific readouts. To enable consistent evaluation across tasks, we introduce distributional metrics that compare distributional properties directly in the space of downstream tasks and apply this framework to nine models and four tasks. Our results highlight the value of bridging representation learning and generative modeling for temporally grounded video understanding.
Authors: Yitong Li, Raoul Grasman
Abstract: The modern recommender systems are facing an increasing challenge of modelling and predicting the dynamic and context-rich user preferences. Traditional collaborative filtering and content-based methods often struggle to capture the temporal patternings and evolving user intentions. While Large Language Models (LLMs) have gained gradual attention in recent years, by their strong semantic understanding and reasoning abilities, they are not inherently designed to model chronologically evolving user preference and intentions. On the other hand, for sequential models like LSTM (Long-Short-Term-Memory) which is good at capturing the temporal dynamics of user behaviour and evolving user preference over time, but still lacks a rich semantic understanding for comprehensive recommendation generation. In this study, we propose DUALRec (Dynamic User-Aware Language-based Recommender), a novel recommender that leverages the complementary strength of both models, which combines the temporal modelling abilities of LSTM networks with semantic reasoning power of the fine-tuned Large Language Models. The LSTM component will capture users evolving preference through their viewing history, while the fine-tuned LLM variants will leverage these temporal user insights to generate next movies that users might enjoy. Experimental results on MovieLens-1M dataset shows that the DUALRec model outperforms a wide range of baseline models, with comprehensive evaluation matrices of Hit Rate (HR@k), Normalized Discounted Cumulative Gain (NDCG@k), and genre similarity metrics. This research proposes a novel architecture that bridges the gap between temporal sequence modeling and semantic reasoning, and offers a promising direction for developing more intelligent and context-aware recommenders.
Authors: Haoyu He, Haozheng Luo, Yan Chen, Qi R. Wang
Abstract: We introduce RHYTHM (Reasoning with Hierarchical Temporal Tokenization for Human Mobility), a framework that leverages large language models (LLMs) as spatio-temporal predictors and trajectory reasoners. RHYTHM partitions trajectories into daily segments encoded as discrete tokens with hierarchical attention, capturing both daily and weekly dependencies while substantially reducing the sequence length. Token representations are enriched with pre-computed prompt embeddings via a frozen LLM, enhancing the model's ability to capture interdependencies without extensive computational overhead. By freezing the LLM backbone, RHYTHM achieves significant computational efficiency. Evaluation on three real-world datasets demonstrates a 2.4% improvement in accuracy, 5.0% increase on weekends, and 24.6% reduction in training time compared to state-of-the-art methods.
Authors: Jianfei Li, Kevin Kam Fung Yuen
Abstract: This study proposes the Cognitive Pairwise Comparison Classification Model Selection (CPC-CMS) framework for document-level sentiment analysis. The CPC, based on expert knowledge judgment, is used to calculate the weights of evaluation criteria, including accuracy, precision, recall, F1-score, specificity, Matthews Correlation Coefficient (MCC), Cohen's Kappa (Kappa), and efficiency. Naive Bayes, Linear Support Vector Classification (LSVC), Random Forest, Logistic Regression, Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and A Lite Bidirectional Encoder Representations from Transformers (ALBERT) are chosen as classification baseline models. A weighted decision matrix consisting of classification evaluation scores with respect to criteria weights, is formed to select the best classification model for a classification problem. Three open datasets of social media are used to demonstrate the feasibility of the proposed CPC-CMS. Based on our simulation, for evaluation results excluding the time factor, ALBERT is the best for the three datasets; if time consumption is included, no single model always performs better than the other models. The CPC-CMS can be applied to the other classification applications in different areas.
Authors: Zhanli Wu, Fabrizio Leisen, F. Javier Rubio
Abstract: Regression problems with bounded continuous outcomes frequently arise in real-world statistical and machine learning applications, such as the analysis of rates and proportions. A central challenge in this setting is predicting a response associated with a new covariate value. Most of the existing statistical and machine learning literature has focused either on point prediction of bounded outcomes or on interval prediction based on asymptotic approximations. We develop conformal prediction intervals for bounded outcomes based on transformation models and beta regression. We introduce tailored non-conformity measures based on residuals that are aligned with the underlying models, and account for the inherent heteroscedasticity in regression settings with bounded outcomes. We present a theoretical result on asymptotic marginal and conditional validity in the context of full conformal prediction, which remains valid under model misspecification. For split conformal prediction, we provide an empirical coverage analysis based on a comprehensive simulation study. The simulation study demonstrates that both methods provide valid finite-sample predictive coverage, including settings with model misspecification. Finally, we demonstrate the practical performance of the proposed conformal prediction intervals on real data and compare them with bootstrap-based alternatives.
Authors: Hao Fang, Sihao Teng, Hao Yu, Siyi Yuan, Huaiwu He, Zhe Liu, Yunjie Yang
Abstract: Electrical Impedance Tomography (EIT) is a non-invasive, low-cost bedside imaging modality with high temporal resolution, making it suitable for bedside monitoring. However, its inherently ill-posed inverse problem poses significant challenges for accurate image reconstruction. Deep learning (DL)-based approaches have shown promise but often rely on complex network architectures with a large number of parameters, limiting efficiency and scalability. Here, we propose an Ultra-Lightweight Quantum-Assisted Inference (QuantEIT) framework for EIT image reconstruction. QuantEIT leverages a Quantum-Assisted Network (QA-Net), combining parallel 2-qubit quantum circuits to generate expressive latent representations that serve as implicit nonlinear priors, followed by a single linear layer for conductivity reconstruction. This design drastically reduces model complexity and parameter number. Uniquely, QuantEIT operates in an unsupervised, training-data-free manner and represents the first integration of quantum circuits into EIT image reconstruction. Extensive experiments on simulated and real-world 2D and 3D EIT lung imaging data demonstrate that QuantEIT outperforms conventional methods, achieving comparable or superior reconstruction accuracy using only 0.2% of the parameters, with enhanced robustness to noise.
Authors: Hao Fang, Hao Yu, Sihao Teng, Tao Zhang, Siyi Yuan, Huaiwu He, Zhe Liu, Yunjie Yang
Abstract: Unsupervised learning methods, such as Deep Image Prior (DIP), have shown great potential in tomographic imaging due to their training-data-free nature and high generalization capability. However, their reliance on numerous network parameter iterations results in high computational costs, limiting their practical application, particularly in complex 3D or time-sequence tomographic imaging tasks. To overcome these challenges, we propose Deep Dynamic Image Prior (D2IP), a novel framework for 3D time-sequence imaging. D2IP introduces three key strategies - Unsupervised Parameter Warm-Start (UPWS), Temporal Parameter Propagation (TPP), and a customized lightweight reconstruction backbone, 3D-FastResUNet - to accelerate convergence, enforce temporal coherence, and improve computational efficiency. Experimental results on both simulated and clinical pulmonary datasets demonstrate that D2IP enables fast and accurate 3D time-sequence Electrical Impedance Tomography (tsEIT) reconstruction. Compared to state-of-the-art baselines, D2IP delivers superior image quality, with a 24.8% increase in average MSSIM and an 8.1% reduction in ERR, alongside significantly reduced computational time (7.1x faster), highlighting its promise for clinical dynamic pulmonary imaging.
Authors: Marcel Hedman, Desi R. Ivanova, Cong Guan, Tom Rainforth
Abstract: We develop a semi-amortized, policy-based, approach to Bayesian experimental design (BED) called Stepwise Deep Adaptive Design (Step-DAD). Like existing, fully amortized, policy-based BED approaches, Step-DAD trains a design policy upfront before the experiment. However, rather than keeping this policy fixed, Step-DAD periodically updates it as data is gathered, refining it to the particular experimental instance. This test-time adaptation improves both the flexibility and the robustness of the design strategy compared with existing approaches. Empirically, Step-DAD consistently demonstrates superior decision-making and robustness compared with current state-of-the-art BED methods.
Authors: Temiloluwa Prioleau, Baiying Lu, Yanjun Cui
Abstract: Artificial intelligence (AI) algorithms are a critical part of state-of-the-art digital health technology for diabetes management. Yet, access to large high-quality datasets is creating barriers that impede development of robust AI solutions. To accelerate development of transparent, reproducible, and robust AI solutions, we present Glucose-ML, a collection of 10 publicly available diabetes datasets, released within the last 7 years (i.e., 2018 - 2025). The Glucose-ML collection comprises over 300,000 days of continuous glucose monitor (CGM) data with a total of 38 million glucose samples collected from 2500+ people across 4 countries. Participants include persons living with type 1 diabetes, type 2 diabetes, prediabetes, and no diabetes. To support researchers and innovators with using this rich collection of diabetes datasets, we present a comparative analysis to guide algorithm developers with data selection. Additionally, we conduct a case study for the task of blood glucose prediction - one of the most common AI tasks within the field. Through this case study, we provide a benchmark for short-term blood glucose prediction across all 10 publicly available diabetes datasets within the Glucose-ML collection. We show that the same algorithm can have significantly different prediction results when developed/evaluated with different datasets. Findings from this study are then used to inform recommendations for developing robust AI solutions within the diabetes or broader health domain. We provide direct links to each longitudinal diabetes dataset in the Glucose-ML collection and openly provide our code.
Authors: Garapati Keerthana, Manik Gupta
Abstract: Progress notes are among the most clinically meaningful artifacts in an Electronic Health Record (EHR), offering temporally grounded insights into a patient's evolving condition, treatments, and care decisions. Despite their importance, they are severely underrepresented in large-scale EHR datasets. For instance, in the widely used Medical Information Mart for Intensive Care III (MIMIC-III) dataset, only about $8.56\%$ of hospital visits include progress notes, leaving gaps in longitudinal patient narratives. In contrast, the dataset contains a diverse array of other note types, each capturing different aspects of care. We present DENSE (Documenting Evolving Progress Notes from Scattered Evidence), a system designed to align with clinical documentation workflows by simulating how physicians reference past encounters while drafting progress notes. The system introduces a fine-grained note categorization and a temporal alignment mechanism that organizes heterogeneous notes across visits into structured, chronological inputs. At its core, DENSE leverages a clinically informed retrieval strategy to identify temporally and semantically relevant content from both current and prior visits. This retrieved evidence is used to prompt a large language model (LLM) to generate clinically coherent and temporally aware progress notes. We evaluate DENSE on a curated cohort of patients with multiple visits and complete progress note documentation. The generated notes demonstrate strong longitudinal fidelity, achieving a temporal alignment ratio of $1.089$, surpassing the continuity observed in original notes. By restoring narrative coherence across fragmented documentation, our system supports improved downstream tasks such as summarization, predictive modeling, and clinical decision support, offering a scalable solution for LLM-driven note synthesis in real-world healthcare settings.
Authors: \v{S}imon Kubov, Simon Kl\'i\v{c}n\'ik, Jakub Dand\'ar, Zden\v{e}k Straka, Karol\'ina Kvakov\'a, Daniel Kvak
Abstract: Scoliosis affects roughly 2 to 4 percent of adolescents, and treatment decisions depend on precise Cobb angle measurement. Manual assessment is time consuming and subject to inter observer variation. We conducted a retrospective, multi centre evaluation of a fully automated deep learning software (Carebot AI Bones, Spine Measurement functionality; Carebot s.r.o.) on 103 standing anteroposterior whole spine radiographs collected from ten hospitals. Two musculoskeletal radiologists independently measured each study and served as reference readers. Agreement between the AI and each radiologist was assessed with Bland Altman analysis, mean absolute error (MAE), root mean squared error (RMSE), Pearson correlation coefficient, and Cohen kappa for four grade severity classification. Against Radiologist 1 the AI achieved an MAE of 3.89 degrees (RMSE 4.77 degrees) with a bias of 0.70 degrees and limits of agreement from minus 8.59 to plus 9.99 degrees. Against Radiologist 2 the AI achieved an MAE of 3.90 degrees (RMSE 5.68 degrees) with a bias of 2.14 degrees and limits from minus 8.23 to plus 12.50 degrees. Pearson correlations were r equals 0.906 and r equals 0.880 (inter reader r equals 0.928), while Cohen kappa for severity grading reached 0.51 and 0.64 (inter reader kappa 0.59). These results demonstrate that the proposed software reproduces expert level Cobb angle measurements and categorical grading across multiple centres, suggesting its utility for streamlining scoliosis reporting and triage in clinical workflows.
Authors: Shravan Venkatraman, Pavan Kumar S, Rakesh Raj Madavan, Chandrakala S
Abstract: Accurate classification of computed tomography (CT) images is essential for diagnosis and treatment planning, but existing methods often struggle with the subtle and spatially diverse nature of pathological features. Current approaches typically process images uniformly, limiting their ability to detect localized abnormalities that require focused analysis. We introduce UGPL, an uncertainty-guided progressive learning framework that performs a global-to-local analysis by first identifying regions of diagnostic ambiguity and then conducting detailed examination of these critical areas. Our approach employs evidential deep learning to quantify predictive uncertainty, guiding the extraction of informative patches through a non-maximum suppression mechanism that maintains spatial diversity. This progressive refinement strategy, combined with an adaptive fusion mechanism, enables UGPL to integrate both contextual information and fine-grained details. Experiments across three CT datasets demonstrate that UGPL consistently outperforms state-of-the-art methods, achieving improvements of 3.29%, 2.46%, and 8.08% in accuracy for kidney abnormality, lung cancer, and COVID-19 detection, respectively. Our analysis shows that the uncertainty-guided component provides substantial benefits, with performance dramatically increasing when the full progressive learning pipeline is implemented. Our code is available at: https://github.com/shravan-18/UGPL
Authors: Xinyu Cao, Bimal Adhikari, Shangqing Zhao, Jingxian Wu, Yanjun Pan
Abstract: Radio frequency (RF) fingerprinting, which extracts unique hardware imperfections of radio devices, has emerged as a promising physical-layer device identification mechanism in zero trust architectures and beyond 5G networks. In particular, deep learning (DL) methods have demonstrated state-of-the-art performance in this domain. However, existing approaches have primarily focused on enhancing system robustness against temporal and spatial variations in wireless environments, while the security vulnerabilities of these DL-based approaches have often been overlooked. In this work, we systematically investigate the security risks of DL-based RF fingerprinting systems through an adversarial-driven experimental analysis. We observe a consistent misclassification behavior for DL models under domain shifts, where a device is frequently misclassified as another specific one. Our analysis based on extensive real-world experiments demonstrates that this behavior can be exploited as an effective backdoor to enable external attackers to intrude into the system. Furthermore, we show that training DL models on raw received signals causes the models to entangle RF fingerprints with environmental and signal-pattern features, creating additional attack vectors that cannot be mitigated solely through post-processing security methods such as confidence thresholds.
Authors: Xiaoya Li, Xiaofei Sun, Albert Wang, Jiwei Li, Chris Shum
Abstract: The exponential growth in demand for GPU computing resources, driven by the rapid advancement of Large Language Models, has created an urgent need for automated CUDA optimization strategies. While recent advances in LLMs show promise for code generation, current SOTA models (e.g. R1, o1) achieve low success rates in improving CUDA speed. In this paper, we introduce CUDA-L1, an automated reinforcement learning framework for CUDA optimization. CUDA-L1 achieves performance improvements on the CUDA optimization task: trained on NVIDIA A100, it delivers an average speedup of x17.7 across all 250 CUDA kernels of KernelBench, with peak speedups reaching x449. Furthermore, the model also demonstrates excellent portability across GPU architectures, achieving average speedups of x17.8 on H100, x19.0 on RTX 3090, x16.5 on L40, x14.7 on H800, and x13.9 on H20 despite being optimized specifically for A100. Beyond these benchmark results, CUDA-L1 demonstrates several remarkable properties: 1) Discovers a variety of CUDA optimization techniques and learns to combine them strategically to achieve optimal performance; 2) Uncovers fundamental principles of CUDA optimization; 3) Identifies non-obvious performance bottlenecks and rejects seemingly beneficial optimizations that harm performance. The capabilities of CUDA-L1 demonstrate that reinforcement learning can transform an initially poor-performing LLM into an effective CUDA optimizer through speedup-based reward signals alone, without human expertise or domain knowledge. More importantly, the trained RL model extend the acquired reasoning abilities to new kernels. This paradigm opens possibilities for automated optimization of CUDA operations, and holds promise to substantially promote GPU efficiency and alleviate the rising pressure on GPU computing resources.
Authors: Dani\"elle Schuman, Mark V. Seebode, Tobias Rohe, Maximilian Balthasar Mansky, Michael Schroedl-Baumann, Jonas Stein, Claudia Linnhoff-Popien, Florian Krellner
Abstract: Exploiting the fact that samples drawn from a quantum annealer inherently follow a Boltzmann-like distribution, annealing-based Quantum Boltzmann Machines (QBMs) have gained increasing popularity in the quantum research community. While they harbor great promises for quantum speed-up, their usage currently stays a costly endeavor, as large amounts of QPU time are required to train them. This limits their applicability in the NISQ era. Following the idea of No\`e et al. (2024), who tried to alleviate this cost by incorporating parallel quantum annealing into their unsupervised training of QBMs, this paper presents an improved version of parallel quantum annealing that we employ to train QBMs in a supervised setting. Saving qubits to encode the inputs, the latter setting allows us to test our approach on medical images from the MedMNIST data set (Yang et al., 2023), thereby moving closer to real-world applicability of the technology. Our experiments show that QBMs using our approach already achieve reasonable results, comparable to those of similarly-sized Convolutional Neural Networks (CNNs), with markedly smaller numbers of epochs than these classical models. Our parallel annealing technique leads to a speed-up of almost 70 % compared to regular annealing-based BM executions.
Authors: Maksim Kuprashevich, Grigorii Alekseenko, Irina Tolstykh, Georgii Fedorov, Bulat Suleimanov, Vladimir Dokholyan, Aleksandr Gordeev
Abstract: Recent advances in generative modeling enable image editing assistants that follow natural language instructions without additional user input. Their supervised training requires millions of triplets: original image, instruction, edited image. Yet mining pixel-accurate examples is hard. Each edit must affect only prompt-specified regions, preserve stylistic coherence, respect physical plausibility, and retain visual appeal. The lack of robust automated edit-quality metrics hinders reliable automation at scale. We present an automated, modular pipeline that mines high-fidelity triplets across domains, resolutions, instruction complexities, and styles. Built on public generative models and running without human intervention, our system uses a task-tuned Gemini validator to score instruction adherence and aesthetics directly, removing any need for segmentation or grounding models. Inversion and compositional bootstrapping enlarge the mined set by approximately 2.2x, enabling large-scale high-fidelity training data. By automating the most repetitive annotation steps, the approach allows a new scale of training without human labeling effort. To democratize research in this resource-intensive area, we release NHR-Edit: an open dataset of 358k high-quality triplets. In the largest cross-dataset evaluation, it surpasses all public alternatives. We also release Bagel-NHR-Edit, an open-source fine-tuned Bagel model, which achieves state-of-the-art metrics in our experiments.
Authors: Subrata Biswas, Mohammad Nur Hossain Khan, Violet Colwell, Jack Adiletta, Bashima Islam
Abstract: Accurate sound source localization (SSL), such as direction-of-arrival (DoA) estimation, relies on consistent multichannel data. However, batteryless systems often suffer from missing data due to the stochastic nature of energy harvesting, degrading localization performance. We propose LOCUS, a deep learning framework that recovers corrupted features in such settings. LOCUS integrates three modules: (1) Information-Weighted Focus (InFo) to identify corrupted regions, (2) Latent Feature Synthesizer (LaFS) to reconstruct missing features, and (3) Guided Replacement (GRep) to restore data without altering valid inputs. LOCUS significantly improves DoA accuracy under missing-channel conditions, achieving up to 36.91% error reduction on DCASE and LargeSet, and 25.87-59.46% gains in real-world deployments. We release a 50-hour multichannel dataset to support future research on localization under energy constraints. Our code and data are available at: https://bashlab.github.io/locus_project/
Authors: Kevin Chang, Nathan Dahlin, Rahul Jain, Pierluigi Nuzzo
Abstract: Over the past decade, neural network (NN)-based controllers have demonstrated remarkable efficacy in a variety of decision-making tasks. However, their black-box nature and the risk of unexpected behaviors pose a challenge to their deployment in real-world systems requiring strong guarantees of correctness and safety. We address these limitations by investigating the transformation of NN-based controllers into equivalent soft decision tree (SDT)-based controllers and its impact on verifiability. In contrast to existing work, we focus on discrete-output NN controllers including rectified linear unit (ReLU) activation functions as well as argmax operations. We then devise an exact yet efficient transformation algorithm which automatically prunes redundant branches. We first demonstrate the practical efficacy of the transformation algorithm applied to an autonomous driving NN controller within OpenAI Gym's CarRacing environment. Subsequently, we evaluate our approach using two benchmarks from the OpenAI Gym environment. Our results indicate that the SDT transformation can benefit formal verification, showing runtime improvements of up to $21 \times$ and $2 \times$ for MountainCar-v0 and CartPole-v1, respectively.
Authors: Wenliang Liu, Danyang Li, Erfan Aasi, Daniela Rus, Roberto Tron, Calin Belta
Abstract: Imitation learning methods have demonstrated considerable success in teaching autonomous systems complex tasks through expert demonstrations. However, a limitation of these methods is their lack of interpretability, particularly in understanding the specific task the learning agent aims to accomplish. In this paper, we propose a novel imitation learning method that combines Signal Temporal Logic (STL) inference and control synthesis, enabling the explicit representation of the task as an STL formula. This approach not only provides a clear understanding of the task but also supports the integration of human knowledge and allows for adaptation to out-of-distribution scenarios by manually adjusting the STL formulas and fine-tuning the policy. We employ a Generative Adversarial Network (GAN)-inspired approach to train both the inference and policy networks, effectively narrowing the gap between expert and learned policies. The efficiency of our algorithm is demonstrated through simulations, showcasing its practical applicability and adaptability.
Authors: Arturs Berzins, Andreas Radler, Eric Volkmann, Sebastian Sanokowski, Sepp Hochreiter, Johannes Brandstetter
Abstract: Geometry is a ubiquitous tool in computer graphics, design, and engineering. However, the lack of large shape datasets limits the application of state-of-the-art supervised learning methods and motivates the exploration of alternative learning strategies. To this end, we introduce geometry-informed neural networks (GINNs) -- a framework for training shape-generative neural fields without data by leveraging user-specified design requirements in the form of objectives and constraints. By adding diversity as an explicit constraint, GINNs avoid mode-collapse and can generate multiple diverse solutions, often required in geometry tasks. Experimentally, we apply GINNs to several problems spanning physics, geometry, and engineering design, showing control over geometrical and topological properties, such as surface smoothness or the number of holes. These results demonstrate the potential of training shape-generative models without data, paving the way for new generative design approaches without large datasets.
Authors: Simon Letzgus, Klaus-Robert M\"uller, Gr\'egoire Montavon
Abstract: In recent years, Explainable AI (XAI) methods have facilitated profound validation and knowledge extraction from ML models. While extensively studied for classification, few XAI solutions have addressed the challenges specific to regression models. In regression, explanations need to be precisely formulated to address specific user queries (e.g.\ distinguishing between `Why is the output above 0?' and `Why is the output above 50?'). They should furthermore reflect the model's behavior on the relevant data sub-manifold. In this paper, we introduce XpertAI, a framework that disentangles the prediction strategy into multiple range-specific sub-strategies and allows the formulation of precise queries about the model (the `explanandum') as a linear combination of those sub-strategies. XpertAI is formulated generally to work alongside popular XAI attribution techniques, based on occlusion, gradient integration, or reverse propagation. Qualitative and quantitative results, demonstrate the benefits of our approach.
Authors: Jon Vadillo, Roberto Santana, Jose A. Lozano, Marta Kwiatkowska
Abstract: The lack of transparency of Deep Neural Networks continues to be a limitation that severely undermines their reliability and usage in high-stakes applications. Promising approaches to overcome such limitations are Prototype-Based Self-Explainable Neural Networks (PSENNs), whose predictions rely on the similarity between the input at hand and a set of prototypical representations of the output classes, offering therefore a deep, yet transparent-by-design, architecture. In this paper, we introduce a probabilistic reformulation of PSENNs, called Prob-PSENN, which replaces point estimates for the prototypes with probability distributions over their values. This provides not only a more flexible framework for an end-to-end learning of prototypes, but can also capture the explanatory uncertainty of the model, which is a missing feature in previous approaches. In addition, since the prototypes determine both the explanation and the prediction, Prob-PSENNs allow us to detect when the model is making uninformed or uncertain predictions, and to obtain valid explanations for them. Our experiments demonstrate that Prob-PSENNs provide more meaningful and robust explanations than their non-probabilistic counterparts, while remaining competitive in terms of predictive performance, thus enhancing the explainability and reliability of the models.
Authors: Thom Badings, Wietze Koops, Sebastian Junges, Nils Jansen
Abstract: We consider the verification of neural network policies for discrete-time stochastic systems with respect to reach-avoid specifications. We use a learner-verifier procedure that learns a certificate for the specification, represented as a neural network. Verifying that this neural network certificate is a so-called reach-avoid supermartingale (RASM) proves the satisfaction of a reach-avoid specification. Existing approaches for such a verification task rely on computed Lipschitz constants of neural networks. These approaches struggle with large Lipschitz constants, especially for reach-avoid specifications with high threshold probabilities. We present two key contributions to obtain smaller Lipschitz constants than existing approaches. First, we introduce logarithmic RASMs (logRASMs), which take exponentially smaller values than RASMs and hence have lower theoretical Lipschitz constants. Second, we present a fast method to compute tighter upper bounds on Lipschitz constants based on weighted norms. Our empirical evaluation shows we can consistently verify the satisfaction of reach-avoid specifications with probabilities as high as 99.9999%.
Authors: Ties Robroek (IT University of Copenhagen), Neil Kim Nielsen (IT University of Copenhagen), P{\i}nar T\"oz\"un (IT University of Copenhagen)
Abstract: Training deep learning models is a repetitive and resource-intensive process. Data scientists often train several models before landing on a set of parameters (e.g., hyper-parameter tuning) and model architecture (e.g., neural architecture search), among other things that yield the highest accuracy. The computational efficiency of these training tasks depends highly on how well the training data is supplied to the training process. The repetitive nature of these tasks results in the same data processing pipelines running over and over, exacerbating the need for and costs of computational resources. In this paper, we present TensorSocket to reduce the computational needs of deep learning training by enabling simultaneous training processes to share the same data loader. TensorSocket mitigates CPU-side bottlenecks in cases where the collocated training workloads have high throughput on GPU, but are held back by lower data-loading throughput on CPU. TensorSocket achieves this by reducing redundant computations and data duplication across collocated training processes and leveraging modern GPU-GPU interconnects. While doing so, TensorSocket is able to train and balance differently-sized models and serve multiple batch sizes simultaneously and is hardware- and pipeline-agnostic in nature. Our evaluation shows that TensorSocket enables scenarios that are infeasible without data sharing, increases training throughput by up to 100%, and when utilizing cloud instances, achieves cost savings of 50% by reducing the hardware resource needs on the CPU side. Furthermore, TensorSocket outperforms the state-of-the-art solutions for shared data loading such as CoorDL and Joader; it is easier to deploy and maintain and either achieves higher or matches their throughput while requiring fewer CPU resources.
Authors: Brandon Knutson, Amandin Chyba Rabeendran, Michael Ivanitskiy, Jordan Pettyjohn, Cecilia Diniz-Behn, Samy Wu Fung, Daniel McKenzie
Abstract: Recent work suggests that certain neural network architectures -- particularly recurrent neural networks (RNNs) and implicit neural networks (INNs) -- are capable of logical extrapolation. When trained on easy instances of a task, these networks (henceforth: logical extrapolators) can generalize to more difficult instances. Previous research has hypothesized that logical extrapolators do so by learning a scalable, iterative algorithm for the given task which converges to the solution. We examine this idea more closely in the context of a single task: maze solving. By varying test data along multiple axes -- not just maze size -- we show that models introduced in prior work fail in a variety of ways, some expected and others less so. It remains uncertain whether any of these models has truly learned an algorithm. However, we provide evidence that a certain RNN has approximately learned a form of `deadend-filling'. We show that training these models on more diverse data addresses some failure modes but, paradoxically, does not improve logical extrapolation. We also analyze convergence behavior, and show that models explicitly trained to converge to a fixed point are likely to do so when extrapolating, while models that are not may exhibit more exotic limiting behavior such as limit cycles, even when they correctly solve the problem. Our results (i) show that logical extrapolation is not immune to the problem of goal misgeneralization, and (ii) suggest that analyzing the dynamics of extrapolation may yield insights into designing better logical extrapolators.
Authors: Yan-Ru Ju, Tai-Lin Wu, Chung-Chin Shih, Ti-Rong Wu
Abstract: Although AlphaZero has achieved superhuman performance in board games, recent studies reveal its limitations in handling scenarios requiring a comprehensive understanding of the entire board, such as recognizing long-sequence patterns in Go. To address this challenge, we propose ResTNet, a network that interleaves residual and Transformer blocks to bridge local and global knowledge. ResTNet improves playing strength across multiple board games, increasing win rate from 54.6% to 60.8% in 9x9 Go, 53.6% to 60.9% in 19x19 Go, and 50.4% to 58.0% in 19x19 Hex. In addition, ResTNet effectively processes global information and tackles two long-sequence patterns in 19x19 Go, including circular pattern and ladder pattern. It reduces the mean square error for circular pattern recognition from 2.58 to 1.07 and lowers the attack probability against an adversary program from 70.44% to 23.91%. ResTNet also improves ladder pattern recognition accuracy from 59.15% to 80.01%. By visualizing attention maps, we demonstrate that ResTNet captures critical game concepts in both Go and Hex, offering insights into AlphaZero's decision-making process. Overall, ResTNet shows a promising approach to integrating local and global knowledge, paving the way for more effective AlphaZero-based algorithms in board games. Our code is available at https://rlg.iis.sinica.edu.tw/papers/restnet.
Authors: Ruikai Yang, Mingzhen He, Zhengbao He, Youmei Qiu, Xiaolin Huang
Abstract: Machine unlearning (MU) is to make a well-trained model behave as if it had never been trained on specific data. In today's over-parameterized models, dominated by neural networks, a common approach is to manually relabel data and fine-tune the well-trained model. It can approximate the MU model in the output space, but the question remains whether it can achieve exact MU, i.e., in the parameter space. We answer this question by employing random feature techniques to construct an analytical framework. Under the premise of model optimization via stochastic gradient descent, we theoretically demonstrated that over-parameterized linear models can achieve exact MU through relabeling specific data. We also extend this work to real-world nonlinear networks and propose an alternating optimization algorithm that unifies the tasks of unlearning and relabeling. The algorithm's effectiveness, confirmed through numerical experiments, highlights its superior performance in unlearning across various scenarios compared to current state-of-the-art methods, particularly excelling over similar relabeling-based MU approaches.
Authors: Kevin Tirta Wijaya, Minghao Guo, Michael Sun, Hans-Peter Seidel, Wojciech Matusik, Vahid Babaei
Abstract: Molecular deep learning models have achieved remarkable success in property prediction, but they often require large amounts of labeled data. The challenge is that, in real-world applications, labels are extremely scarce, as obtaining them through laboratory experimentation is both expensive and time-consuming. In this work, we introduce MoleVers, a versatile pretrained molecular model designed for various types of molecular property prediction in the wild, i.e., where experimentally-validated labels are scarce. MoleVers employs a two-stage pretraining strategy. In the first stage, it learns molecular representations from unlabeled data through masked atom prediction and extreme denoising, a novel task enabled by our newly introduced branching encoder architecture and dynamic noise scale sampling. In the second stage, the model refines these representations through predictions of auxiliary properties derived from computational methods, such as the density functional theory or large language models. Evaluation on 22 small, experimentally-validated datasets demonstrates that MoleVers achieves state-of-the-art performance, highlighting the effectiveness of its two-stage framework in producing generalizable molecular representations for diverse downstream properties.
Authors: Hao Chen, Han Tao, Guo Song, Jie Zhang, Yunlong Yu, Yonghan Dong, Lei Bai
Abstract: This paper presents Variables Adaptive Mixture of Experts (VAMoE), a novel framework for incremental weather forecasting that dynamically adapts to evolving spatiotemporal patterns in real time data. Traditional weather prediction models often struggle with exorbitant computational expenditure and the need to continuously update forecasts as new observations arrive. VAMoE addresses these challenges by leveraging a hybrid architecture of experts, where each expert specializes in capturing distinct subpatterns of atmospheric variables (temperature, humidity, wind speed). Moreover, the proposed method employs a variable adaptive gating mechanism to dynamically select and combine relevant experts based on the input context, enabling efficient knowledge distillation and parameter sharing. This design significantly reduces computational overhead while maintaining high forecast accuracy. Experiments on real world ERA5 dataset demonstrate that VAMoE performs comparable against SoTA models in both short term (1 days) and long term (5 days) forecasting tasks, with only about 25% of trainable parameters and 50% of the initial training data.
Authors: Jiang Yang, Yuxiang Zhao, Quanhui Zhu
Abstract: Understanding the training dynamics of deep neural networks (DNNs), particularly how they evolve low-dimensional features from high-dimensional data, remains a central challenge in deep learning theory. In this work, we introduce the concept of $\epsilon$-rank, a novel metric quantifying the effective feature of neuron functions in the terminal hidden layer. Through extensive experiments across diverse tasks, we observe a universal staircase phenomenon: during training process implemented by the standard stochastic gradient descent methods, the decline of the loss function is accompanied by an increase in the $\epsilon$-rank and exhibits a staircase pattern. Theoretically, we rigorously prove a negative correlation between the loss lower bound and $\epsilon$-rank, demonstrating that a high $\epsilon$-rank is essential for significant loss reduction. Moreover, numerical evidences show that within the same deep neural network, the $\epsilon$-rank of the subsequent hidden layer is higher than that of the previous hidden layer. Based on these observations, to eliminate the staircase phenomenon, we propose a novel pre-training strategy on the initial hidden layer that elevates the $\epsilon$-rank of the terminal hidden layer. Numerical experiments validate its effectiveness in reducing training time and improving accuracy across various tasks. Therefore, the newly introduced concept of $\epsilon$-rank is a computable quantity that serves as an intrinsic effective metric characteristic for deep neural networks, providing a novel perspective for understanding the training dynamics of neural networks and offering a theoretical foundation for designing efficient training strategies in practical applications.
Authors: Sunwoong Yang, Ricardo Vinuesa, Namwoo Kang
Abstract: This study addresses the critical challenge of error accumulation in spatio-temporal auto-regressive (AR) predictions within scientific machine learning models by exploring temporal integration schemes and adaptive multi-step rollout strategies. We introduce the first implementation of the two-step Adams-Bashforth method specifically tailored for data-driven AR prediction, leveraging historical derivative information to enhance numerical stability without additional computational overhead. To validate our approach, we systematically evaluate time integration schemes across canonical 2D PDEs before extending to complex Navier-Stokes cylinder vortex shedding dynamics. Additionally, we develop three novel adaptive weighting strategies that dynamically adjust the importance of different future time steps during multi-step rollout training. Our analysis reveals that as physical complexity increases, such sophisticated rollout techniques become essential, with the Adams-Bashforth scheme demonstrating consistent robustness across investigated systems and our best adaptive approach delivering an 89% improvement over conventional fixed-weight methods while maintaining similar computational costs. For the complex Navier-Stokes vortex shedding problem, despite using an extremely lightweight graph neural network with just 1,177 trainable parameters and training on only 50 snapshots, our framework accurately predicts 350 future time steps reducing mean squared error from 0.125 (single-step direct prediction) to 0.002 (Adams-Bashforth with proposed multi-step rollout). Our integrated methodology demonstrates an 83% improvement over standard noise injection techniques and maintains robustness under severe spatial constraints; specifically, when trained on only a partial spatial domain, it still achieves 58% and 27% improvements over direct prediction and forward Euler methods, respectively.
Authors: Konstantin Donhauser, Kristina Ulicna, Gemma Elyse Moran, Aditya Ravuri, Kian Kenyon-Dean, Cian Eastwood, Jason Hartford
Abstract: Sparse dictionary learning (DL) has emerged as a powerful approach to extract semantically meaningful concepts from the internals of large language models (LLMs) trained mainly in the text domain. In this work, we explore whether DL can extract meaningful concepts from less human-interpretable scientific data, such as vision foundation models trained on cell microscopy images, where limited prior knowledge exists about which high-level concepts should arise. We propose a novel combination of a sparse DL algorithm, Iterative Codebook Feature Learning (ICFL), with a PCA whitening pre-processing step derived from control data. Using this combined approach, we successfully retrieve biologically meaningful concepts, such as cell types and genetic perturbations. Moreover, we demonstrate how our method reveals subtle morphological changes arising from human-interpretable interventions, offering a promising new direction for scientific discovery via mechanistic interpretability in bioimaging.
Authors: Di Yu, Xin Du, Linshan Jiang, Huijing Zhang, Shuiguang Deng
Abstract: The energy efficiency of deep spiking neural networks (SNNs) aligns with the constraints of resource-limited edge devices, positioning SNNs as a promising foundation for intelligent applications leveraging the extensive data collected by these devices. To address data privacy concerns when deploying SNNs on edge devices, federated learning (FL) facilitates collaborative model training by leveraging data distributed across edge devices without transmitting local data to a central server. However, existing FL approaches struggle with label-skewed data across devices, which leads to drift in local SNN models and degrades the performance of the global SNN model. In this paper, we propose a novel framework called FedLEC, which incorporates intra-client label weight calibration to balance the learning intensity across local labels and inter-client knowledge distillation to mitigate local SNN model bias caused by label absence. Extensive experiments with three different structured SNNs across five datasets (i.e., three non-neuromorphic and two neuromorphic datasets) demonstrate the efficiency of FedLEC. Compared to eight state-of-the-art FL algorithms, FedLEC achieves an average accuracy improvement of approximately 11.59% for the global SNN model under various label skew distribution settings.
Authors: Mathias Bellat, Jordy D. Orellana Figueroa, Jonathan S. Reeves, Ruhollah Taghizadeh-Mehrjardi, Claudio Tennie, Thomas Scholten
Abstract: Artificial intelligence and machine learning applications in archaeology have increased significantly in recent years, and these now span all subfields, geographical regions, and time periods. The prevalence and success of these applications have remained largely unexamined, as recent reviews on the use of machine learning in archaeology have only focused only on specific subfields of archaeology. Our review examined an exhaustive corpus of 135 articles published between 1997 and 2022. We observed a significant increase in the number of publications from 2019 onwards. Automatic structure detection and artefact classification were the most represented tasks in the articles reviewed, followed by taphonomy, and archaeological predictive modelling. From the review, clustering and unsupervised methods were underrepresented compared to supervised models. Artificial neural networks and ensemble learning account for two thirds of the total number of models used. However, if machine learning models are gaining in popularity they remain subject to misunderstanding. We observed, in some cases, poorly defined requirements and caveats of the machine learning methods used. Furthermore, the goals and the needs of machine learning applications for archaeological purposes are in some cases unclear or poorly expressed. To address this, we proposed a workflow guide for archaeologists to develop coherent and consistent methodologies adapted to their research questions, project scale and data. As in many other areas, machine learning is rapidly becoming an important tool in archaeological research and practice, useful for the analyses of large and multivariate data, although not without limitations. This review highlights the importance of well-defined and well-reported structured methodologies and collaborative practices to maximise the potential of applications of machine learning methods in archaeology.
Authors: Lukas Moosbrugger, Valentin Seiler, Philipp Wohlgenannt, Sebastian Hegenbart, Sashko Ristov, Elias Eder, Peter Kepplinger
Abstract: Energy communities (ECs) play a key role in enabling local demand shifting and enhancing self-sufficiency, as energy systems transition toward decentralized structures with high shares of renewable generation. To optimally operate them, accurate short-term load forecasting is essential, particularly for implementing demand-side management strategies. With the recent rise of deep learning methods, data-driven forecasting has gained significant attention, however, it remains insufficiently explored in many practical contexts. Therefore, this study evaluates the effectiveness of state-of-the-art deep learning models-including LSTM, xLSTM, and Transformer architectures-compared to traditional benchmarks such as K-Nearest Neighbors (KNN) and persistence forecasting, across varying community size, historical data availability, and model complexity. Additionally, we assess the benefits of transfer learning using publicly available synthetic load profiles. On average, transfer learning improves the normalized mean absolute error by 1.97 percentage points when only two months of training data are available. Interestingly, for less than six months of training data, simple persistence models outperform deep learning architectures in forecast accuracy. The practical value of improved forecasting is demonstrated using a mixed-integer linear programming optimization for ECs with a shared battery energy storage system. For an energy community with 50 households, the most accurate deep learning model achieves an average reduction in financial energy costs of 8.06%. Notably, a simple KNN approach achieves average savings of 8.01%, making it a competitive and robust alternative. All implementations are publicly available to facilitate reproducibility. These findings offer actionable insights for ECs, and they highlight when the additional complexity of deep learning is warranted by performance gains.
Authors: Raghav Singhal, Zachary Horvitz, Ryan Teehan, Mengye Ren, Zhou Yu, Kathleen McKeown, Rajesh Ranganath
Abstract: Diffusion models produce impressive results in modalities ranging from images and video to protein design and text. However, generating samples with user-specified properties remains a challenge. Recent research proposes fine-tuning models to maximize rewards that capture desired properties, but these methods require expensive training and are prone to mode collapse. In this work, we present Feynman-Kac (FK) steering, an inference-time framework for steering diffusion models with reward functions. FK steering works by sampling a system of multiple interacting diffusion processes, called particles, and resampling particles at intermediate steps based on scores computed using functions called potentials. Potentials are defined using rewards for intermediate states and are selected such that a high value indicates that the particle will yield a high-reward sample. We explore various choices of potentials, intermediate rewards, and samplers. We evaluate FK steering on text-to-image and text diffusion models. For steering text-to-image models with a human preference reward, we find that FK steering a 0.8B parameter model outperforms a 2.6B parameter fine-tuned model on prompt fidelity, with faster sampling and no training. For steering text diffusion models with rewards for text quality and specific text attributes, we find that FK steering generates lower perplexity, more linguistically acceptable outputs and enables gradient-free control of attributes like toxicity. Our results demonstrate that inference-time scaling and steering of diffusion models - even with off-the-shelf rewards - can provide significant sample quality gains and controllability benefits. Code is available at https://github.com/zacharyhorvitz/Fk-Diffusion-Steering .
URLs: https://github.com/zacharyhorvitz/Fk-Diffusion-Steering
Authors: Qitao Tan, Jun Liu, Zheng Zhan, Caiwei Ding, Yanzhi Wang, Xiaolong Ma, Jaewoo Lee, Jin Lu, Geng Yuan
Abstract: Large language models (LLMs) excel across various tasks, but standard first-order (FO) fine-tuning demands considerable memory, significantly limiting real-world deployment. Recently, zeroth-order (ZO) optimization stood out as a promising memory-efficient training paradigm, avoiding backward passes and relying solely on forward passes for gradient estimation, making it attractive for resource-constrained scenarios. However, ZO method lags far behind FO method in both convergence speed and accuracy. To bridge the gap, we introduce a novel layer-wise divergence analysis that uncovers the distinct update pattern of FO and ZO optimization. Aiming to resemble the learning capacity of FO method from the findings, we propose Divergence-driven Zeroth-Order (DiZO) optimization. DiZO conducts divergence-driven layer adaptation by incorporating projections to ZO updates, generating diverse-magnitude updates precisely scaled to layer-wise individual optimization needs. Our results demonstrate that DiZO significantly reduces the needed iterations for convergence without sacrificing throughput, cutting training GPU hours by up to 48% on various datasets. Moreover, DiZO consistently outperforms the representative ZO baselines in fine-tuning RoBERTa-large, OPT-series, and Llama-series on downstream tasks and, in some cases, even surpasses memory-intensive FO fine-tuning. Our code is released at https://anonymous.4open.science/r/DiZO-E86D.
Authors: Juan Du, Dongheng Chen
Abstract: Anomaly detection based on 3D point cloud data is an important research problem and receives more and more attention recently. Untrained anomaly detection based on only one sample is an emerging research problem motivated by real manufacturing industries such as personalized manufacturing where only one sample can be collected without any additional labels and historical datasets. Identifying anomalies accurately based on one 3D point cloud sample is a critical challenge in both industrial applications and the field of machine learning. This paper aims to provide a formal definition of the untrained anomaly detection problem based on 3D point cloud data, discuss the differences between untrained anomaly detection and current unsupervised anomaly detection problems. Unlike trained unsupervised learning, untrained unsupervised learning does not rely on any data, including unlabeled data. Instead, they leverage prior knowledge about the surfaces and anomalies. We propose three complementary methodological frameworks: the Latent Variable Inference Framework that employs probabilistic modeling to distinguish anomalies; the Decomposition Framework that separates point clouds into reference, anomaly, and noise components through sparse learning; and the Local Geometry Framework that leverages neighborhood information for anomaly identification. Experimental results demonstrate that untrained methods achieve competitive detection performance while offering significant computational advantages, demonstrating up to a 15-fold increase in execution speed. The proposed methods provide viable solutions for scenarios with extreme data scarcity, addressing critical challenges in personalized manufacturing and healthcare applications where collecting multiple samples or historical data is infeasible.
Authors: Finn Rietz, Oleg Smirnov, Sara Karimi, Lele Cao
Abstract: Prompting has emerged as the dominant paradigm for adapting large, pre-trained transformer-based models to downstream tasks. The Prompting Decision Transformer (PDT) enables large-scale, multi-task offline Reinforcement Learning (RL) pre-training by leveraging stochastic trajectory prompts to identify the target task. However, these prompts are sampled uniformly from expert demonstrations, overlooking a critical limitation: not all prompts are equally informative for differentiating between tasks. This limits generalization and adaptation, especially in low-data or open-world settings where sample efficiency is crucial. To address this issue, we propose a lightweight, inference-time, bandit-based prompt-tuning framework. The bandit explores and optimizes trajectory prompt selection to enhance task performance, while avoiding costly fine-tuning of the transformer backbone. Our experiments indicate not only clear performance gains due to bandit-based prompt-tuning, but also better sample complexity, scalability, and prompt space exploration compared to prompt-tuning baselines. These results highlights the importance of adaptive prompt selection mechanisms for efficient generalization in offline multi-task RL.
Authors: Thomas Foster, Anya Sims, Johannes Forkel, Mattie Fellows, Jakob Foerster
Abstract: Reinforcement learning is now widely adopted as the final stage of large language model training, especially for reasoning-style tasks such as maths problems. Typically, models attempt each question many times during a single training step and attempt to learn from their successes and failures. However, we demonstrate that throughout training with two popular algorithms (PPO and VinePPO) on two widely used datasets, many questions are either solved by all attempts - meaning they are already learned - or by none - providing no meaningful training signal. To address this, we adapt a method from the reinforcement learning literature - sampling for learnability - and apply it to the reinforcement learning stage of LLM training. Our curriculum prioritises questions with high variance of success, i.e. those where the agent sometimes succeeds, but not always. Our findings demonstrate that this curriculum consistently boosts training performance across multiple algorithms and datasets, paving the way for more efficient and effective reinforcement learning with LLMs.
Authors: Haoyuan Wu, Haisheng Zheng, Shoubo Hu, Zhuolun He, Bei Yu
Abstract: Logic synthesis, a critical stage in electronic design automation (EDA), optimizes gate-level circuits to minimize power consumption and area occupancy in integrated circuits (ICs). Traditional logic synthesis tools rely on human-designed heuristics, often yielding suboptimal results. Although differentiable architecture search (DAS) has shown promise in generating circuits from truth tables, it faces challenges such as high computational complexity, convergence to local optima, and extensive hyperparameter tuning. Consequently, we propose a novel approach integrating conditional generative models with DAS for circuit generation. Our approach first introduces CircuitVQ, a circuit tokenizer trained based on our Circuit AutoEncoder We then develop CircuitAR, a masked autoregressive model leveraging CircuitVQ as the tokenizer. CircuitAR can generate preliminary circuit structures from truth tables, which guide DAS in producing functionally equivalent circuits. Notably, we observe the scalability and emergent capability in generating complex circuit structures of our CircuitAR models. Extensive experiments also show the superior performance of our method. This research bridges the gap between probabilistic generative models and precise circuit generation, offering a robust solution for logic synthesis.
Authors: Enoch H. Kang, Hema Yoganarasimhan, Lalit Jain
Abstract: We study the problem of estimating Dynamic Discrete Choice (DDC) models, also known as offline Maximum Entropy-Regularized Inverse Reinforcement Learning (offline MaxEnt-IRL) in machine learning. The objective is to recover reward or $Q^*$ functions that govern agent behavior from offline behavior data. In this paper, we propose a globally convergent gradient-based method for solving these problems without the restrictive assumption of linearly parameterized rewards. The novelty of our approach lies in introducing the Empirical Risk Minimization (ERM) based IRL/DDC framework, which circumvents the need for explicit state transition probability estimation in the Bellman equation. Furthermore, our method is compatible with non-parametric estimation techniques such as neural networks. Therefore, the proposed method has the potential to be scaled to high-dimensional, infinite state spaces. A key theoretical insight underlying our approach is that the Bellman residual satisfies the Polyak-Lojasiewicz (PL) condition -- a property that, while weaker than strong convexity, is sufficient to ensure fast global convergence guarantees. Through a series of synthetic experiments, we demonstrate that our approach consistently outperforms benchmark methods and state-of-the-art alternatives.
Authors: Bum Geun Park, Taeho Lee, Donghwan Lee
Abstract: This paper introduces Q-learning with gradient target tracking, a novel reinforcement learning framework that provides a learned continuous target update mechanism as an alternative to the conventional hard update paradigm. In the standard deep Q-network (DQN), the target network is a copy of the online network's weights, held fixed for a number of iterations before being periodically replaced via a hard update. While this stabilizes training by providing consistent targets, it introduces a new challenge: the hard update period must be carefully tuned to achieve optimal performance. To address this issue, we propose two gradient-based target update methods: DQN with asymmetric gradient target tracking (AGT2-DQN) and DQN with symmetric gradient target tracking (SGT2-DQN). These methods replace the conventional hard target updates with continuous and structured updates using gradient descent, which effectively eliminates the need for manual tuning. We provide a theoretical analysis proving the convergence of these methods in tabular settings. Additionally, empirical evaluations demonstrate their advantages over standard DQN baselines, which suggest that gradient-based target updates can serve as an effective alternative to conventional target update mechanisms in Q-learning.
Authors: Srinitish Srinivasan, Omkumar CU
Abstract: While hyperbolic GNNs show promise for hierarchical data, they often have limited discriminative power compared to Euclidean counterparts or the WL test, due to non-injective aggregation. To address this expressivity gap, we propose the Lorentzian Graph Isomorphic Network (LGIN), a novel HGNN designed for enhanced discrimination within the Lorentzian model. LGIN introduces a new update rule that preserves the Lorentzian metric while effectively capturing richer structural information. This marks a significant step towards more expressive GNNs on Riemannian manifolds. Extensive evaluations across nine benchmark datasets demonstrate LGIN's superior performance, consistently outperforming or matching state-of-the-art hyperbolic and Euclidean baselines, showcasing its ability to capture complex graph structures. LGIN is the first to adapt principles of powerful, highly discriminative GNN architectures to a Riemannian manifold. The code for our paper can be found at https://github.com/Deceptrax123/LGIN
Authors: Kiran K. Yalamanchi, Pinaki Pal, Balaji Mohan, Abdullah S. AlRamadan, Jihad A. Badra, Yuanjiang Pei
Abstract: In the present work, a generative deep learning framework combining a Co-optimized Variational Autoencoder (Co-VAE) architecture with quantitative structure-property relationship (QSPR) techniques is developed to enable accelerated inverse design of fuels. The Co-VAE integrates a property prediction component coupled with the VAE latent space, enhancing molecular reconstruction and accurate estimation of Research Octane Number (RON) (chosen as the fuel property of interest). A subset of the GDB-13 database, enriched with a curated RON database, is used for model training. Hyperparameter tuning is further utilized to optimize the balance among reconstruction fidelity, chemical validity, and RON prediction. An independent regression model is then used to refine RON prediction, while a differential evolution algorithm is employed to efficiently navigate the VAE latent space and identify promising fuel molecule candidates with high RON. This methodology addresses the limitations of traditional fuel screening approaches by capturing complex structure-property relationships within a comprehensive latent representation. The generative model can be adapted to different target properties, enabling systematic exploration of large chemical spaces relevant to fuel design applications. Furthermore, the demonstrated framework can be readily extended by incorporating additional synthesizability criteria to improve applicability and reliability for de novo design of new fuels.
Authors: Tamim Al Mahmud, Najeeb Jebreel, Josep Domingo-Ferrer, David Sanchez
Abstract: Large language models (LLMs) have recently revolutionized language processing tasks but have also brought ethical and legal issues. LLMs have a tendency to memorize potentially private or copyrighted information present in the training data, which might then be delivered to end users at inference time. When this happens, a naive solution is to retrain the model from scratch after excluding the undesired data. Although this guarantees that the target data have been forgotten, it is also prohibitively expensive for LLMs. Approximate unlearning offers a more efficient alternative, as it consists of ex post modifications of the trained model itself to prevent undesirable results, but it lacks forgetting guarantees because it relies solely on empirical evidence. In this work, we present DP2Unlearning, a novel LLM unlearning framework that offers formal forgetting guarantees at a significantly lower cost than retraining from scratch on the data to be retained. DP2Unlearning involves training LLMs on textual data protected using {\epsilon}-differential privacy (DP), which later enables efficient unlearning with the guarantees against disclosure associated with the chosen {\epsilon}. Our experiments demonstrate that DP2Unlearning achieves similar model performance post-unlearning, compared to an LLM retraining from scratch on retained data -- the gold standard exact unlearning -- but at approximately half the unlearning cost. In addition, with a reasonable computational cost, it outperforms approximate unlearning methods at both preserving the utility of the model post-unlearning and effectively forgetting the targeted information.
Authors: James Giroux, Cristiano Fanelli
Abstract: We present a (proto) Foundation Model for Nuclear Physics, capable of operating on low-level detector inputs from Imaging Cherenkov Detectors at the future Electron Ion Collider. Building upon established next-token prediction approaches, we aim to address potential challenges such as resolution loss from existing tokenization schemes and limited support for conditional generation. We propose four key innovations: (i) separate vocabularies for discrete and continuous variates, combined via Causal Multi-Head Cross-Attention (CMHCA), (ii) continuous kinematic conditioning through prepended context embeddings, (iii) scalable and simple, high-resolution continuous variate tokenization without joint vocabulary inflation, and (iv) class conditional generation through a Mixture of Experts. Our model enables fast, high-fidelity generation of pixel and time sequences for Cherenkov photons, validated through closure tests in the High Performance DIRC. We also show our model generalizes to reconstruction tasks such as pion/kaon identification, and noise filtering, in which we show its ability to leverage fine-tuning under specific objectives.
Authors: Dirk Tasche
Abstract: Recalibration of binary probabilistic classifiers to a target prior probability is an important task in areas like credit risk management. We analyse methods for recalibration from a distribution shift perspective. Distribution shift assumptions linked to the area under the curve (AUC) of a probabilistic classifier are found to be useful for the design of meaningful recalibration methods. Two new methods called parametric covariate shift with posterior drift (CSPD) and ROC-based quasi moment matching (QMM) are proposed and tested together with some other methods in an example setting. The outcomes of the test suggest that the QMM methods discussed in the paper can provide appropriately conservative results in evaluations with concave functionals like for instance risk weights functions for credit risk.
Authors: Daehyeon Baek, Jieun Choi, Jimyoung Son, Kyungmin Bin, Seungbeom Choi, Kihyo Moon, Minsung Jang, Hyojung Lee
Abstract: As large language models become increasingly prevalent, memory bandwidth constraints significantly limit inference throughput, motivating post-training quantization (PTQ). In this paper, we propose FireQ, a co-designed PTQ framework and an INT4-FP8 matrix multiplication kernel that accelerates LLM inference across all linear layers. Specifically, FireQ quantizes linear layer weights and key-values to INT4, and activations and queries to FP8, significantly enhancing throughput. Additionally, we introduce a three-stage pipelining for the prefill phase, which modifies the FlashAttention-3 kernel, effectively reducing time-to-first-token in the prefill phase. To minimize accuracy loss from quantization, we develop novel outlier smoothing techniques tailored separately for linear and attention layers. In linear layers, we explicitly use per-tensor scaling to prevent underflow caused by the FP8 quantization scaling factor of INT4 quantization, and channel-wise scaling to compensate for coarse granularity of INT4. In attention layers, we address quantization challenges posed by rotary positional embeddings (RoPE) by combining pre-RoPE and post-RoPE scaling strategies. FireQ significantly outperforms state-of-the-art methods, achieving 1.68x faster inference in feed-forward network layers on Llama2-7B and 1.26x faster prefill phase performance on Llama3-8B compared to QServe, with negligible accuracy loss.
Authors: Jacob Piland, Chris Sweet, Adam Czajka
Abstract: Class Activation Mapping (CAM) and its gradient-based variants (e.g., GradCAM) have become standard tools for explaining Convolutional Neural Network (CNN) predictions. However, these approaches typically focus on individual logits, while for neural networks using softmax, the class membership probability estimates depend \textit{only} on the \textit{differences} between logits, not on their absolute values. This disconnect leaves standard CAMs vulnerable to adversarial manipulation, such as passive fooling, where a model is trained to produce misleading CAMs without affecting decision performance. We introduce \textbf{Salience-Hoax Activation Maps (SHAMs)}, an \emph{entropy-aware form of passive fooling} that serves as a benchmark for CAM robustness under adversarial conditions. To address the passive fooling vulnerability, we then propose \textbf{DiffGradCAM}, a novel, lightweight, and contrastive approach to class activation mapping that is both non-suceptible to passive fooling, but also matches the output of standard CAM methods such as GradCAM in the non-adversarial case. Together, SHAM and DiffGradCAM establish a new framework for probing and improving the robustness of saliency-based explanations. We validate both contributions across multi-class tasks with few and many classes.
Authors: Xiaowen Ma, Chenyang Lin, Yao Zhang, Volker Tresp, Yunpu Ma
Abstract: Leveraging multiple Large Language Models(LLMs) has proven effective for addressing complex, high-dimensional tasks, but current approaches often rely on static, manually engineered multi-agent configurations. To overcome these constraints, we present the Agentic Neural Network(ANN), a framework that conceptualizes multi-agent collaboration as a layered neural network architecture. In this design, each agent operates as a node, and each layer forms a cooperative "team" focused on a specific subtask. Agentic Neural Network follows a two-phase optimization strategy: (1) Forward Phase-Drawing inspiration from neural network forward passes, tasks are dynamically decomposed into subtasks, and cooperative agent teams with suitable aggregation methods are constructed layer by layer. (2) Backward Phase-Mirroring backpropagation, we refine both global and local collaboration through iterative feedback, allowing agents to self-evolve their roles, prompts, and coordination. This neuro-symbolic approach enables ANN to create new or specialized agent teams post-training, delivering notable gains in accuracy and adaptability. Across four benchmark datasets, ANN surpasses leading multi-agent baselines under the same configurations, showing consistent performance improvements. Our findings indicate that ANN provides a scalable, data-driven framework for multi-agent systems, combining the collaborative capabilities of LLMs with the efficiency and flexibility of neural network principles. We plan to open-source the entire framework.
Authors: Kondrup Emma
Abstract: Dynamic link prediction remains a central challenge in temporal graph learning, particularly in designing models that are both effective and practical for real-world deployment. Existing approaches often rely on complex neural architectures, which are computationally intensive and difficult to interpret. In this work, we build on the strong recurrence-based foundation of the EdgeBank baseline, by supplementing it with inductive capabilities. We do so by leveraging the predictive power of non-learnable signals from two complementary perspectives: historical edge recurrence, as captured by EdgeBank, and global node popularity, as introduced in the PopTrack model. We propose t-CoMem, a lightweight memory module that tracks temporal co-occurrence patterns and neighborhood activity. Building on this, we introduce Base3, an interpolation-based model that fuses EdgeBank, PopTrack, and t-CoMem into a unified scoring framework. This combination effectively bridges local and global temporal dynamics -- repetition, popularity, and context -- without relying on training. Evaluated on the Temporal Graph Benchmark, Base3 achieves performance competitive with state-of-the-art deep models, even outperforming them on some datasets. Importantly, it considerably improves on existing baselines' performance under more realistic and challenging negative sampling strategies -- offering a simple yet robust alternative for temporal graph learning.
Authors: Yanfang Hou, Carlos Fern\'andez-Lor\'ia
Abstract: Causal forests estimate how treatment effects vary across individuals, guiding personalized interventions in areas like marketing, operations, and public policy. A standard modeling practice with this method is honest estimation: dividing the data so that the subgroups used to model treatment effect variation are formed separately from the data used to estimate those effects. This is intended to reduce overfitting and is the default in many software packages. But is it always the right choice? In this paper, we show that honest estimation can reduce the accuracy of individual-level treatment effect estimates, especially when there are substantial differences in how individuals respond to treatment, and the data is rich enough to uncover those differences. The core issue is a classic bias-variance trade-off: honesty lowers the risk of overfitting but increases the risk of underfitting, because it limits the data available to detect patterns. Across 7,500 benchmark datasets, we find that the cost of using honesty by default can be as high as requiring 75% more data to match the performance of models trained without it. We argue that honesty is best understood as a form of regularization, and like any regularization choice, its use should be guided by out-of-sample performance, not adopted reflexively.
Authors: Han Liu, Keyan Ding, Peilin Chen, Yinwei Wei, Liqiang Nie, Dapeng Wu, Shiqi Wang
Abstract: Accurate prediction of protein-ligand binding affinity is critical for drug discovery. While recent deep learning approaches have demonstrated promising results, they often rely solely on structural features of proteins and ligands, overlooking their valuable biochemical knowledge associated with binding affinity. To address this limitation, we propose KEPLA, a novel deep learning framework that explicitly integrates prior knowledge from Gene Ontology and ligand properties to enhance prediction performance. KEPLA takes protein sequences and ligand molecular graphs as input and optimizes two complementary objectives: (1) aligning global representations with knowledge graph relations to capture domain-specific biochemical insights, and (2) leveraging cross attention between local representations to construct fine-grained joint embeddings for prediction. Experiments on two benchmark datasets across both in-domain and cross-domain scenarios demonstrate that KEPLA consistently outperforms state-of-the-art baselines. Furthermore, interpretability analyses based on knowledge graph relations and cross attention maps provide valuable insights into the underlying predictive mechanisms.
Authors: Constantin Venhoff, Iv\'an Arcuschin, Philip Torr, Arthur Conmy, Neel Nanda
Abstract: Recent advances in large language models (LLMs) have led to the development of thinking language models that generate extensive internal reasoning chains before producing responses. While these models achieve improved performance, controlling their reasoning processes remains challenging. This work presents a steering approach for thinking LLMs by analyzing and manipulating specific reasoning behaviors in DeepSeek-R1-Distill models. Through a systematic experiment on 500 tasks across 10 diverse categories, we identify several reasoning behaviors exhibited by thinking models, including expressing uncertainty, generating examples for hypothesis validation, and backtracking in reasoning chains. We demonstrate that these behaviors are mediated by linear directions in the model's activation space and can be controlled using steering vectors. By extracting and applying these vectors, we provide a method to modulate specific aspects of the model's reasoning process, such as its tendency to backtrack or express uncertainty. Our approach offers practical tools for steering reasoning processes in thinking models in a controlled and interpretable manner. We validate our steering method using three DeepSeek-R1-Distill models, demonstrating consistent control across different model architectures.
Authors: Chenhao Si, Ming Yan
Abstract: Physics-informed neural networks (PINNs) are extensively employed to solve partial differential equations (PDEs) by ensuring that the outputs and gradients of deep learning models adhere to the governing equations. However, constrained by computational limitations, PINNs are typically optimized using a finite set of points, which poses significant challenges in guaranteeing their convergence and accuracy. In this study, we proposed a new weighting scheme that will adaptively change the weights to the loss functions from isolated points to their continuous neighborhood regions. The empirical results show that our weighting scheme can reduce the relative $L^2$ errors to a lower value.
Authors: Karim Abdel Sadek, Matthew Farrugia-Roberts, Usman Anwar, Hannah Erlebach, Christian Schroeder de Witt, David Krueger, Michael Dennis
Abstract: Safe generalization in reinforcement learning requires not only that a learned policy acts capably in new situations, but also that it uses its capabilities towards the pursuit of the designer's intended goal. The latter requirement may fail when a proxy goal incentivizes similar behavior to the intended goal within the training environment, but not in novel deployment environments. This creates the risk that policies will behave as if in pursuit of the proxy goal, rather than the intended goal, in deployment -- a phenomenon known as goal misgeneralization. In this paper, we formalize this problem setting in order to theoretically study the possibility of goal misgeneralization under different training objectives. We show that goal misgeneralization is possible under approximate optimization of the maximum expected value (MEV) objective, but not the minimax expected regret (MMER) objective. We then empirically show that the standard MEV-based training method of domain randomization exhibits goal misgeneralization in procedurally-generated grid-world environments, whereas current regret-based unsupervised environment design (UED) methods are more robust to goal misgeneralization (though they don't find MMER policies in all cases). Our findings suggest that minimax expected regret is a promising approach to mitigating goal misgeneralization.
Authors: Eric Xing, Mingkai Deng, Jinyu Hou, Zhiting Hu
Abstract: World Model, the supposed algorithmic surrogate of the real-world environment which biological agents experience with and act upon, has been an emerging topic in recent years because of the rising needs to develop virtual agents with artificial (general) intelligence. There has been much debate on what a world model really is, how to build it, how to use it, and how to evaluate it. In this essay, starting from the imagination in the famed Sci-Fi classic Dune, and drawing inspiration from the concept of "hypothetical thinking" in psychology literature, we offer critiques of several schools of thoughts on world modeling, and argue the primary goal of a world model to be simulating all actionable possibilities of the real world for purposeful reasoning and acting. Building on the critiques, we propose a new architecture for a general-purpose world model, based on hierarchical, multi-level, and mixed continuous/discrete representations, and a generative and self-supervision learning framework, with an outlook of a Physical, Agentic, and Nested (PAN) AGI system enabled by such a model.
Authors: Huan Wang, Haoran Li, Huaming Chen, Jun Yan, Jiahua Shi, Jun Shen
Abstract: Federated learning aims at training models collaboratively across participants while protecting privacy. However, one major challenge for this paradigm is the data heterogeneity issue, where biased data preferences across multiple clients, harming the model's convergence and performance. In this paper, we first introduce powerful diffusion models into the federated learning paradigm and show that diffusion representations are effective steers during federated training. To explore the possibility of using diffusion representations in handling data heterogeneity, we propose a novel diffusion-inspired Federated paradigm with Diffusion Representation Collaboration, termed FedDifRC, leveraging meaningful guidance of diffusion models to mitigate data heterogeneity. The key idea is to construct text-driven diffusion contrasting and noise-driven diffusion regularization, aiming to provide abundant class-related semantic information and consistent convergence signals. On the one hand, we exploit the conditional feedback from the diffusion model for different text prompts to build a text-driven contrastive learning strategy. On the other hand, we introduce a noise-driven consistency regularization to align local instances with diffusion denoising representations, constraining the optimization region in the feature space. In addition, FedDifRC can be extended to a self-supervised scheme without relying on any labeled data. We also provide a theoretical analysis for FedDifRC to ensure convergence under non-convex objectives. The experiments on different scenarios validate the effectiveness of FedDifRC and the efficiency of crucial components.
Authors: Burak Demirel, Yu Wang, Cristian Tatino, Pablo Soldati
Abstract: Modern RAN operate in highly dynamic and heterogeneous environments, where hand-tuned, rule-based RRM algorithms often underperform. While RL can surpass such heuristics in constrained settings, the diversity of deployments and unpredictable radio conditions introduce major generalization challenges. Data-driven policies frequently overfit to training conditions, degrading performance in unseen scenarios. To address this, we propose a generalization-centered RL framework for RAN control that: (i) robustly reconstructs dynamically varying states from partial and noisy observations, while encoding static and semi-static information, such as radio nodes, cell attributes, and their topology, through graph representations; (ii) applies domain randomization to broaden the training distribution; and (iii) distributes data generation across multiple actors while centralizing training in a cloud-compatible architecture aligned with O-RAN principles. Although generalization increases computational and data-management complexity, our distributed design mitigates this by scaling data collection and training across diverse network conditions. Applied to downlink link adaptation in five 5G benchmarks, our policy improves average throughput and spectral efficiency by ~10% over an OLLA baseline (10% BLER target) in full-buffer MIMO/mMIMO and by >20% under high mobility. It matches specialized RL in full-buffer traffic and achieves up to 4- and 2-fold gains in eMBB and mixed-traffic benchmarks, respectively. In nine-cell deployments, GAT models offer 30% higher throughput over MLP baselines. These results, combined with our scalable architecture, offer a path toward AI-native 6G RAN using a single, generalizable RL agent.
Authors: Aakash Tripathi, Ian E. Nielsen, Muhammad Umer, Ravi P. Ramachandran, Ghulam Rasool
Abstract: Transcription Factor Binding Site (TFBS) prediction is crucial for understanding gene regulation and various biological processes. This study introduces a novel Mixture of Experts (MoE) approach for TFBS prediction, integrating multiple pre-trained Convolutional Neural Network (CNN) models, each specializing in different TFBS patterns. We evaluate the performance of our MoE model against individual expert models on both in-distribution and out-of-distribution (OOD) datasets, using six randomly selected transcription factors (TFs) for OOD testing. Our results demonstrate that the MoE model achieves competitive or superior performance across diverse TF binding sites, particularly excelling in OOD scenarios. The Analysis of Variance (ANOVA) statistical test confirms the significance of these performance differences. Additionally, we introduce ShiftSmooth, a novel attribution mapping technique that provides more robust model interpretability by considering small shifts in input sequences. Through comprehensive explainability analysis, we show that ShiftSmooth offers superior attribution for motif discovery and localization compared to traditional Vanilla Gradient methods. Our work presents an efficient, generalizable, and interpretable solution for TFBS prediction, potentially enabling new discoveries in genome biology and advancing our understanding of transcriptional regulation.
Authors: Zhenyuan Chen
Abstract: Inductive bias is a key factor in spatial regression models, determining how well a model can learn from limited data and capture spatial patterns. This work revisits the inductive biases in Geographically Neural Network Weighted Regression (GNNWR) and identifies limitations in current approaches for modeling spatial non-stationarity. While GNNWR extends traditional Geographically Weighted Regression by using neural networks to learn spatial weighting functions, existing implementations are often restricted by fixed distance-based schemes and limited inductive bias. We propose to generalize GNNWR by incorporating concepts from convolutional neural networks, recurrent neural networks, and transformers, introducing local receptive fields, sequential context, and self-attention into spatial regression. Through extensive benchmarking on synthetic spatial datasets with varying heterogeneity, noise, and sample sizes, we show that GNNWR outperforms classic methods in capturing nonlinear and complex spatial relationships. Our results also reveal that model performance depends strongly on data characteristics, with local models excelling in highly heterogeneous or small-sample scenarios, and global models performing better with larger, more homogeneous data. These findings highlight the importance of inductive bias in spatial modeling and suggest future directions, including learnable spatial weighting functions, hybrid neural architectures, and improved interpretability for models handling non-stationary spatial data.
Authors: \'Etienne K\"unzel, Achref Jaziri, Visvanathan Ramesh
Abstract: Continual learning in computer vision requires that models adapt to a continuous stream of tasks without forgetting prior knowledge, yet existing approaches often tip the balance heavily toward either plasticity or stability. We introduce RDBP, a simple, low-overhead baseline that unites two complementary mechanisms: ReLUDown, a lightweight activation modification that preserves feature sensitivity while preventing neuron dormancy, and Decreasing Backpropagation, a biologically inspired gradient-scheduling scheme that progressively shields early layers from catastrophic updates. Evaluated on the Continual ImageNet benchmark, RDBP matches or exceeds the plasticity and stability of state-of-the-art methods while reducing computational cost. RDBP thus provides both a practical solution for real-world continual learning and a clear benchmark against which future continual learning strategies can be measured.
Authors: Daniel Commey, Benjamin Appiah, Griffith S. Klogo, Garth V. Crosby
Abstract: Federated Learning (FL) enables collaborative model training on decentralized data without exposing raw data. However, the evaluation phase in FL may leak sensitive information through shared performance metrics. In this paper, we propose a novel protocol that incorporates Zero-Knowledge Proofs (ZKPs) to enable privacy-preserving and verifiable evaluation for FL. Instead of revealing raw loss values, clients generate a succinct proof asserting that their local loss is below a predefined threshold. Our approach is implemented without reliance on external APIs, using self-contained modules for federated learning simulation, ZKP circuit design, and experimental evaluation on both the MNIST and Human Activity Recognition (HAR) datasets. We focus on a threshold-based proof for a simple Convolutional Neural Network (CNN) model (for MNIST) and a multi-layer perceptron (MLP) model (for HAR), and evaluate the approach in terms of computational overhead, communication cost, and verifiability.
Authors: Abhishek Sriram, Neal Tuffy
Abstract: This paper presents a machine learning-accelerated optimization framework for RF power amplifier design that reduces simulation requirements by 65% while maintaining $\pm0.4$ dBm accuracy for the majority of the modes. The proposed method combines MaxMin Latin Hypercube Sampling with CatBoost gradient boosting to intelligently explore multidimensional parameter spaces. Instead of exhaustively simulating all parameter combinations to achieve target P2dB compression specifications, our approach strategically selects approximately 35% of critical simulation points. The framework processes ADS netlists, executes harmonic balance simulations on the reduced dataset, and trains a CatBoost model to predict P2dB performance across the entire design space. Validation across 15 PA operating modes yields an average $R^2$ of 0.901, with the system ranking parameter combinations by their likelihood of meeting target specifications. The integrated solution delivers 58.24% to 77.78% reduction in simulation time through automated GUI-based workflows, enabling rapid design iterations without compromising accuracy standards required for production RF circuits.
Authors: Boris Bonev, Thorsten Kurth, Ankur Mahesh, Mauro Bisson, Jean Kossaifi, Karthik Kashinath, Anima Anandkumar, William D. Collins, Michael S. Pritchard, Alexander Keller
Abstract: FourCastNet 3 advances global weather modeling by implementing a scalable, geometric machine learning (ML) approach to probabilistic ensemble forecasting. The approach is designed to respect spherical geometry and to accurately model the spatially correlated probabilistic nature of the problem, resulting in stable spectra and realistic dynamics across multiple scales. FourCastNet 3 delivers forecasting accuracy that surpasses leading conventional ensemble models and rivals the best diffusion-based methods, while producing forecasts 8 to 60 times faster than these approaches. In contrast to other ML approaches, FourCastNet 3 demonstrates excellent probabilistic calibration and retains realistic spectra, even at extended lead times of up to 60 days. All of these advances are realized using a purely convolutional neural network architecture tailored for spherical geometry. Scalable and efficient large-scale training on 1024 GPUs and more is enabled by a novel training paradigm for combined model- and data-parallelism, inspired by domain decomposition methods in classical numerical models. Additionally, FourCastNet 3 enables rapid inference on a single GPU, producing a 60-day global forecast at 0.25{\deg}, 6-hourly resolution in under 4 minutes. Its computational efficiency, medium-range probabilistic skill, spectral fidelity, and rollout stability at subseasonal timescales make it a strong candidate for improving meteorological forecasting and early warning systems through large ensemble predictions.
Authors: Luca Stradiotti, Dario Pesenti, Stefano Teso, Jesse Davis
Abstract: Machine Learning predictors are increasingly being employed in high-stakes applications such as credit scoring. Explanations help users unpack the reasons behind their predictions, but are not always "high quality''. That is, end-users may have difficulty interpreting or believing them, which can complicate trust assessment and downstream decision-making. We argue that classifiers should have the option to refuse handling inputs whose predictions cannot be explained properly and introduce a framework for learning to reject low-quality explanations (LtX) in which predictors are equipped with a rejector that evaluates the quality of explanations. In this problem setting, the key challenges are how to properly define and assess explanation quality and how to design a suitable rejector. Focusing on popular attribution techniques, we introduce ULER (User-centric Low-quality Explanation Rejector), which learns a simple rejector from human ratings and per-feature relevance judgments to mirror human judgments of explanation quality. Our experiments show that ULER outperforms both state-of-the-art and explanation-aware learning to reject strategies at LtX on eight classification and regression benchmarks and on a new human-annotated dataset, which we will publicly release to support future research.
Authors: Hongze Tan
Abstract: This paper introduces two novel modifications to the Dynamic sAmpling Policy Optimization (DAPO) algorithm [1], approached from a mixed-policy perspective. Standard policy gradient methods can suffer from instability and sample inefficiency, particularly in sparse reward settings. To address this, we first propose a method that incorporates a pre-trained, stable guiding policy ($\piphi$) to provide off-policy experience, thereby regularizing the training of the target policy ($\pion$). This approach improves training stability and convergence speed by adaptively adjusting the learning step size. Secondly, we extend this idea to re-utilize zero-reward samples, which are often discarded by dynamic sampling strategies like DAPO's. By treating these samples as a distinct batch guided by the expert policy, we further enhance sample efficiency. We provide a theoretical analysis for both methods, demonstrating that their objective functions converge to the optimal solution within the established theoretical framework of reinforcement learning. The proposed mixed-policy framework effectively balances exploration and exploitation, promising more stable and efficient policy optimization.
Authors: Kenza Bouzid, Shruthi Bannur, Felix Meissen, Daniel Coelho de Castro, Anton Schwaighofer, Javier Alvarez-Valle, Stephanie L. Hyland
Abstract: Interpretability can improve the safety, transparency and trust of AI models, which is especially important in healthcare applications where decisions often carry significant consequences. Mechanistic interpretability, particularly through the use of sparse autoencoders (SAEs), offers a promising approach for uncovering human-interpretable features within large transformer-based models. In this study, we apply Matryoshka-SAE to the radiology-specialised multimodal large language model, MAIRA-2, to interpret its internal representations. Using large-scale automated interpretability of the SAE features, we identify a range of clinically relevant concepts - including medical devices (e.g., line and tube placements, pacemaker presence), pathologies such as pleural effusion and cardiomegaly, longitudinal changes and textual features. We further examine the influence of these features on model behaviour through steering, demonstrating directional control over generations with mixed success. Our results reveal practical and methodological challenges, yet they offer initial insights into the internal concepts learned by MAIRA-2 - marking a step toward deeper mechanistic understanding and interpretability of a radiology-adapted multimodal large language model, and paving the way for improved model transparency. We release the trained SAEs and interpretations: https://huggingface.co/microsoft/maira-2-sae.
Authors: Etienne Le Naour, Tahar Nabil, Ghislain Agoua
Abstract: Recent years have witnessed a growing interest for time series foundation models, with a strong emphasis on the forecasting task. Yet, the crucial task of out-of-domain imputation of missing values remains largely underexplored. We propose a first step to fill this gap by leveraging implicit neural representations (INRs). INRs model time series as continuous functions and naturally handle various missing data scenarios and sampling rates. While they have shown strong performance within specific distributions, they struggle under distribution shifts. To address this, we introduce MoTM (Mixture of Timeflow Models), a step toward a foundation model for time series imputation. Building on the idea that a new time series is a mixture of previously seen patterns, MoTM combines a basis of INRs, each trained independently on a distinct family of time series, with a ridge regressor that adapts to the observed context at inference. We demonstrate robust in-domain and out-of-domain generalization across diverse imputation scenarios (e.g., block and pointwise missingness, variable sampling rates), paving the way for adaptable foundation imputation models.
Authors: Zikai Xie, Linjiang Chen
Abstract: Bayesian Optimization (BO) algorithm is a standard tool for black-box optimization problems. The current state-of-the-art BO approach for permutation spaces relies on the Mallows kernel-an $\Omega(n^2)$ representation that explicitly enumerates every pairwise comparison. Inspired by the close relationship between the Mallows kernel and pairwise comparison, we propose a novel framework for generating kernel functions on permutation space based on sorting algorithms. Within this framework, the Mallows kernel can be viewed as a special instance derived from bubble sort. Further, we introduce the \textbf{Merge Kernel} constructed from merge sort, which replaces the quadratic complexity with $\Theta(n\log n)$ to achieve the lowest possible complexity. The resulting feature vector is significantly shorter, can be computed in linearithmic time, yet still efficiently captures meaningful permutation distances. To boost robustness and right-invariance without sacrificing compactness, we further incorporate three lightweight, task-agnostic descriptors: (1) a shift histogram, which aggregates absolute element displacements and supplies a global misplacement signal; (2) a split-pair line, which encodes selected long-range comparisons by aligning elements across the two halves of the whole permutation; and (3) sliding-window motifs, which summarize local order patterns that influence near-neighbor objectives. Our empirical evaluation demonstrates that the proposed kernel consistently outperforms the state-of-the-art Mallows kernel across various permutation optimization benchmarks. Results confirm that the Merge Kernel provides a more compact yet more effective solution for Bayesian optimization in permutation space.
Authors: Efe Bozkir, S\"uleyman \"Ozdel, Mengdi Wang, Brendan David-John, Hong Gao, Kevin Butler, Eakta Jain, Enkelejda Kasneci
Abstract: The latest developments in computer hardware, sensor technologies, and artificial intelligence can make virtual reality (VR) and virtual spaces an important part of human everyday life. Eye tracking offers not only a hands-free way of interaction but also the possibility of a deeper understanding of human visual attention and cognitive processes in VR. Despite these possibilities, eye-tracking data also reveals users' privacy-sensitive attributes when combined with the information about the presented stimulus. To address all these possibilities and potential privacy issues, in this survey, we first cover major works in eye tracking, VR, and privacy areas between 2012 and 2022. While eye tracking in the VR part covers the complete pipeline of eye-tracking methodology from pupil detection and gaze estimation to offline use of the data and analyses, as for privacy and security, we focus on eye-based authentication as well as computational methods to preserve the privacy of individuals and their eye-tracking data in VR. Later, considering all of these, we draw three main directions for the research community by focusing on privacy challenges. In summary, this survey provides an extensive literature review of the utmost possibilities with eye tracking in VR and the privacy implications of those possibilities.
Authors: Prasad Gabbur
Abstract: We propose using a Gaussian Mixture Model (GMM) as reverse transition operator (kernel) within the Denoising Diffusion Implicit Models (DDIM) framework, which is one of the most widely used approaches for accelerated sampling from pre-trained Denoising Diffusion Probabilistic Models (DDPM). Specifically we match the first and second order central moments of the DDPM forward marginals by constraining the parameters of the GMM. We see that moment matching is sufficient to obtain samples with equal or better quality than the original DDIM with Gaussian kernels. We provide experimental results with unconditional models trained on CelebAHQ and FFHQ and class-conditional models trained on ImageNet datasets respectively. Our results suggest that using the GMM kernel leads to significant improvements in the quality of the generated samples when the number of sampling steps is small, as measured by FID and IS metrics. For example on ImageNet 256x256, using 10 sampling steps, we achieve a FID of 6.94 and IS of 207.85 with a GMM kernel compared to 10.15 and 196.73 respectively with a Gaussian kernel.
Authors: Pengcheng Hou, Tao Wang, Daniel Cerkoney, Xiansheng Cai, Zhiyi Li, Youjin Deng, Lei Wang, Kun Chen
Abstract: Quantum field theory (QFT) for interacting many-electron systems is fundamental to condensed matter physics, yet achieving accurate solutions confronts computational challenges in managing the combinatorial complexity of Feynman diagrams, implementing systematic renormalization, and evaluating high-dimensional integrals. We present a unifying framework that integrates QFT computational workflows with an AI-powered technology stack. A cornerstone of this framework is representing Feynman diagrams as computational graphs, which structures the inherent mathematical complexity and facilitates the application of optimized algorithms developed for machine learning and high-performance computing. Consequently, automatic differentiation, native to these graph representations, delivers efficient, fully automated, high-order field-theoretic renormalization procedures. This graph-centric approach also enables sophisticated numerical integration; our neural-network-enhanced Monte Carlo method, accelerated via massively parallel GPU implementation, efficiently evaluates challenging high-dimensional diagrammatic integrals. Applying this framework to the uniform electron gas, we determine the quasiparticle effective mass to a precision significantly surpassing current state-of-the-art simulations. Our work demonstrates the transformative potential of integrating AI-driven computational advances with QFT, opening systematic pathways for solving complex quantum many-body problems across disciplines.
Authors: Elisa Sanchez-Bayona, Rodrigo Agerri
Abstract: Metaphors, although occasionally unperceived, are ubiquitous in our everyday language. Thus, it is crucial for Language Models to be able to grasp the underlying meaning of this kind of figurative language. In this work, we present Meta4XNLI, a novel parallel dataset for the tasks of metaphor detection and interpretation that contains metaphor annotations in both Spanish and English. We investigate language models' metaphor identification and understanding abilities through a series of monolingual and cross-lingual experiments by leveraging our proposed corpus. In order to comprehend how these non-literal expressions affect models' performance, we look over the results and perform an error analysis. Additionally, parallel data offers many potential opportunities to investigate metaphor transferability between these languages and the impact of translation on the development of multilingual annotated resources.
Authors: Jie Wang, March Boedihardjo, Yao Xie
Abstract: Optimal transport has been very successful for various machine learning tasks; however, it is known to suffer from the curse of dimensionality. Hence, dimensionality reduction is desirable when applied to high-dimensional data with low-dimensional structures. The kernel max-sliced (KMS) Wasserstein distance is developed for this purpose by finding an optimal nonlinear mapping that reduces data into $1$ dimension before computing the Wasserstein distance. However, its theoretical properties have not yet been fully developed. In this paper, we provide sharp finite-sample guarantees under milder technical assumptions compared with state-of-the-art for the KMS $p$-Wasserstein distance between two empirical distributions with $n$ samples for general $p\in[1,\infty)$. Algorithm-wise, we show that computing the KMS $2$-Wasserstein distance is NP-hard, and then we further propose a semidefinite relaxation (SDR) formulation (which can be solved efficiently in polynomial time) and provide a relaxation gap for the obtained solution. We provide numerical examples to demonstrate the good performance of our scheme for high-dimensional two-sample testing.
Authors: Yixiao Zhang, Yukara Ikemiya, Woosung Choi, Naoki Murata, Marco A. Mart\'inez-Ram\'irez, Liwei Lin, Gus Xia, Wei-Hsiang Liao, Yuki Mitsufuji, Simon Dixon
Abstract: Recent advances in text-to-music editing, which employ text queries to modify music (e.g.\ by changing its style or adjusting instrumental components), present unique challenges and opportunities for AI-assisted music creation. Previous approaches in this domain have been constrained by the necessity to train specific editing models from scratch, which is both resource-intensive and inefficient; other research uses large language models to predict edited music, resulting in imprecise audio reconstruction. To Combine the strengths and address these limitations, we introduce Instruct-MusicGen, a novel approach that finetunes a pretrained MusicGen model to efficiently follow editing instructions such as adding, removing, or separating stems. Our approach involves a modification of the original MusicGen architecture by incorporating a text fusion module and an audio fusion module, which allow the model to process instruction texts and audio inputs concurrently and yield the desired edited music. Remarkably, Instruct-MusicGen only introduces 8% new parameters to the original MusicGen model and only trains for 5K steps, yet it achieves superior performance across all tasks compared to existing baselines, and demonstrates performance comparable to the models trained for specific tasks. This advancement not only enhances the efficiency of text-to-music editing but also broadens the applicability of music language models in dynamic music production environments.
Authors: Guillaume Rochette, Mathieu Rochat, Matthew J. Vowels
Abstract: psifx is a plug-and-play multi-modal feature extraction toolkit, aiming to facilitate and democratize the use of state-of-the-art machine learning techniques for human sciences research. It is motivated by a need (a) to automate and standardize data annotation processes that typically require expensive, lengthy, and inconsistent human labour; (b) to develop and distribute open-source community-driven psychology research software; and (c) to enable large-scale access and ease of use for non-expert users. The framework contains an array of tools for tasks such as speaker diarization, closed-caption transcription and translation from audio; body, hand, and facial pose estimation and gaze tracking with multi-person tracking from video; and interactive textual feature extraction supported by large language models. The package has been designed with a modular and task-oriented approach, enabling the community to add or update new tools easily. This combination creates new opportunities for in-depth study of real-time behavioral phenomena in psychological and social science research.
Authors: Alexander Michael Rombach, Peter Fettke
Abstract: Extracting key information from documents represents a large portion of business workloads and therefore offers a high potential for efficiency improvements and process automation. With recent advances in Deep Learning, a plethora of Deep Learning based approaches for Key Information Extraction have been proposed under the umbrella term Document Understanding that enable the processing of complex business documents. The goal of this systematic literature review is an in-depth analysis of existing approaches in this domain and the identification of opportunities for further research. To this end, 130 approaches published between 2017 and 2024 are analyzed in this study.
Authors: Yunfei Yang
Abstract: This paper studies the problem of how efficiently functions in the Sobolev spaces $\mathcal{W}^{s,q}([0,1]^d)$ and Besov spaces $\mathcal{B}^s_{q,r}([0,1]^d)$ can be approximated by deep ReLU neural networks with width $W$ and depth $L$, when the error is measured in the $L^p([0,1]^d)$ norm. This problem has been studied by several recent works, which obtained the approximation rate $\mathcal{O}((WL)^{-2s/d})$ up to logarithmic factors when $p=q=\infty$, and the rate $\mathcal{O}(L^{-2s/d})$ for networks with fixed width when the Sobolev embedding condition $1/q -1/p
Authors: Ebube Alor, Ahmad Abdellatif, SayedHassan Khatoonabadi, Emad Shihab
Abstract: Software engineering (SE) chatbots are increasingly gaining attention for their role in enhancing development processes. At the core of chatbots are Natural Language Understanding platforms (NLUs), which enable them to comprehend user queries but require labeled data for training. However, acquiring such labeled data for SE chatbots is challenging due to the scarcity of high-quality datasets, as training requires specialized vocabulary and phrases not found in typical language datasets. Consequently, developers often resort to manually annotating user queries -- a time-consuming and resource-intensive process. Previous approaches require human intervention to generate rules, called labeling functions (LFs), that categorize queries based on specific patterns. To address this issue, we propose an approach to automatically generate LFs by extracting patterns from labeled user queries. We evaluate our approach on four SE datasets and measure performance improvement from training NLUs on queries labeled by the generated LFs. The generated LFs effectively label data with AUC scores up to 85.3% and NLU performance improvements up to 27.2%. Furthermore, our results show that the number of LFs affects labeling performance. We believe that our approach can save time and resources in labeling users' queries, allowing practitioners to focus on core chatbot functionalities rather than manually labeling queries.
Authors: Ernesto Arganda, Marcela Carena, Mart\'in de los Rios, Andres D. Perez, Duncan Rocha, Rosa M. Sand\'a Seoane, Carlos E. M. Wagner
Abstract: The search for weakly interacting matter particles (WIMPs) is one of the main objectives of the High Luminosity Large Hadron Collider (HL-LHC). In this work we use Machine-Learning (ML) techniques to explore WIMP radiative decays into a Dark Matter (DM) candidate in a supersymmetric framework. The minimal supersymmetric WIMP sector includes the lightest neutralino that can provide the observed DM relic density through its co-annihilation with the second lightest neutralino and lightest chargino. Moreover, the direct DM detection cross section rates fulfill current experimental bounds and provide discovery targets for the same region of model parameters in which the radiative decay of the second lightest neutralino into a photon and the lightest neutralino is enhanced. This strongly motivates the search for radiatively decaying neutralinos which, however, suffers from strong backgrounds. We investigate the LHC reach in the search for these radiatively decaying particles by means of cut-based and ML methods and estimate its discovery potential in this well-motivated, new physics scenario. We demonstrate that using ML techniques would enable access to most of the parameter space unexplored by other searches.
Authors: Yingzhen Yang, Ping Li
Abstract: We study nonparametric regression by an over-parameterized two-layer neural network trained by gradient descent (GD) in this paper. We show that, if the neural network is trained by GD with early stopping, then the trained network renders a sharp rate of the nonparametric regression risk of $\mathcal{O}(\epsilon_n^2)$, which is the same rate as that for the classical kernel regression trained by GD with early stopping, where $\epsilon_n$ is the critical population rate of the Neural Tangent Kernel (NTK) associated with the network and $n$ is the size of the training data. It is remarked that our result does not require distributional assumptions about the covariate as long as the covariate is bounded, in a strong contrast with many existing results which rely on specific distributions of the covariates such as the spherical uniform data distribution or distributions satisfying certain restrictive conditions. The rate $\mathcal{O}(\epsilon_n^2)$ is known to be minimax optimal for specific cases, such as the case that the NTK has a polynomial eigenvalue decay rate which happens under certain distributional assumptions on the covariates. Our result formally fills the gap between training a classical kernel regression model and training an over-parameterized but finite-width neural network by GD for nonparametric regression without distributional assumptions on the bounded covariate. We also provide confirmative answers to certain open questions or address particular concerns in the literature of training over-parameterized neural networks by GD with early stopping for nonparametric regression, including the characterization of the stopping time, the lower bound for the network width, and the constant learning rate used in GD.
Authors: Shingo Ayabe, Takuto Otomo, Hiroshi Kera, Kazuhiko Kawamoto
Abstract: Offline reinforcement learning, which learns solely from datasets without environmental interaction, has gained attention. This approach, similar to traditional online deep reinforcement learning, is particularly promising for robot control applications. Nevertheless, its robustness against real-world challenges, such as joint actuator faults in robots, remains a critical concern. This study evaluates the robustness of existing offline reinforcement learning methods using legged robots from OpenAI Gym based on average episodic rewards. For robustness evaluation, we simulate failures by incorporating both random and adversarial perturbations, representing worst-case scenarios, into the joint torque signals. Our experiments show that existing offline reinforcement learning methods exhibit significant vulnerabilities to these action perturbations and are more vulnerable than online reinforcement learning methods, highlighting the need for more robust approaches in this field.
Authors: Jing Fang, Saihao Yan, Xueyu Yin, Yinbo Yu, Chunwei Tian, Jiajia Liu
Abstract: Recent studies have shown that cooperative multi-agent deep reinforcement learning (c-MADRL) is under the threat of backdoor attacks. Once a backdoor trigger is observed, it will perform malicious actions leading to failures or malicious goals. However, existing backdoor attacks suffer from several issues, e.g., instant trigger patterns lack stealthiness, the backdoor is trained or activated by an additional network, or all agents are backdoored. To this end, in this paper, we propose a novel backdoor leverage attack against c-MADRL, BLAST, which attacks the entire multi-agent team by embedding the backdoor only in a single agent. Firstly, we introduce adversary spatiotemporal behavior patterns as the backdoor trigger rather than manual-injected fixed visual patterns or instant status and control the period to perform malicious actions. This method can guarantee the stealthiness and practicality of BLAST. Secondly, we hack the original reward function of the backdoor agent via unilateral guidance to inject BLAST, so as to achieve the \textit{leverage attack effect} that can pry open the entire multi-agent system via a single backdoor agent. We evaluate our BLAST against 3 classic c-MADRL algorithms (VDN, QMIX, and MAPPO) in 2 popular c-MADRL environments (SMAC and Pursuit), and 2 existing defense mechanisms. The experimental results demonstrate that BLAST can achieve a high attack success rate while maintaining a low clean performance variance rate.
Authors: Tom Nuno Wolf, Emre Kavak, Fabian Bongratz, Christian Wachinger
Abstract: The deployment of deep learning models in critical domains necessitates a balance between high accuracy and interpretability. We introduce SIC, an inherently interpretable neural network that provides local and global explanations of its decision-making process. Leveraging the concept of case-based reasoning, SIC extracts class-representative support vectors from training images, ensuring they capture relevant features while suppressing irrelevant ones. Classification decisions are made by calculating and aggregating similarity scores between these support vectors and the input's latent feature vector. We employ B-Cos transformations, which align model weights with inputs, to yield coherent pixel-level explanations in addition to global explanations of case-based reasoning. We evaluate SIC on three tasks: fine-grained classification on Stanford Dogs and FunnyBirds, multi-label classification on Pascal VOC, and pathology detection on the RSNA dataset. Results indicate that SIC not only achieves competitive accuracy compared to state-of-the-art black-box and inherently interpretable models but also offers insightful explanations verified through practical evaluation on the FunnyBirds benchmark. Our theoretical analysis proves that these explanations fulfill established axioms for explanations. Our findings underscore SIC's potential for applications where understanding model decisions is as critical as the decisions themselves.
Authors: Haozhe Wang, Long Li, Chao Qu, Fengming Zhu, Weidi Xu, Wei Chu, Fangzhen Lin
Abstract: Recent advances in mathematical problem-solving with language models (LMs) integrate chain-of-thought (CoT) reasoning and code execution to harness their complementary strengths. However, existing hybrid frameworks exhibit a critical limitation: they depend on externally dictated instructions or rigid code-integration templates, lacking metacognitive awareness -- the capacity to dynamically evaluate intrinsic capabilities and autonomously determine when and how to integrate tools. This rigidity motivates our study of autonomous code integration, enabling models to adapt tool-usage strategies as their reasoning abilities evolve during training. While reinforcement learning (RL) shows promise for boosting LLM reasoning at scale (e.g., DeepSeek-R1), we demonstrate its inefficiency in learning autonomous code integration due to inadequate exploration of the vast combinatorial space of CoT-code interleaving patterns. To address this challenge, we propose a novel Expectation-Maximization (EM) framework that synergizes structured exploration (E-step) with off-policy RL optimization (M-step), creating a self-reinforcing cycle between metacognitive tool-use decisions and evolving capabilities. Experiments reveal our method achieves superior results through improved exploration. Notably, our 7B model improves over 11% on MATH500 and 9.4% on AIME without o1-like CoT.
Authors: Tong Chen, Faeze Brahman, Jiacheng Liu, Niloofar Mireshghallah, Weijia Shi, Pang Wei Koh, Luke Zettlemoyer, Hannaneh Hajishirzi
Abstract: Language models (LMs) can memorize and reproduce segments from their pretraining data verbatim even in non-adversarial settings, raising concerns about copyright, plagiarism, privacy, and creativity. We introduce Paraphrase Preference Optimization (ParaPO), a post-training method that fine-tunes LMs to reduce unintentional regurgitation while preserving their overall utility. ParaPO trains LMs to prefer paraphrased versions of memorized segments over the original verbatim content from the pretraining data. To maintain the ability to recall famous quotations when appropriate, we develop a variant of ParaPO that uses system prompts to control regurgitation behavior. In our evaluation on Llama3.1-8B, ParaPO consistently reduces regurgitation across all tested datasets (e.g., reducing the regurgitation metric from 17.3 to 12.9 in creative writing), whereas unlearning methods used in prior work to mitigate regurgitation are less effective outside their targeted unlearned domain (from 17.3 to 16.9). When applied to the instruction-tuned Tulu3-8B model, ParaPO with system prompting successfully preserves famous quotation recall while reducing unintentional regurgitation (from 8.7 to 6.3 in creative writing) when prompted not to regurgitate. In contrast, without ParaPO tuning, prompting the model not to regurgitate produces only a marginal reduction (8.7 to 8.4).
Authors: Zhiyong Jin, Runhua Xu, Chao Li, Yizhong Liu, Jianxin Li
Abstract: Federated Learning (FL) enables collaborative model training across distributed clients while preserving data privacy, yet it faces significant challenges in communication efficiency and vulnerability to poisoning attacks. While sparsification techniques mitigate communication overhead by transmitting only critical model parameters, they inadvertently amplify security risks: adversarial clients can exploit sparse updates to evade detection and degrade model performance. Existing defense mechanisms, designed for standard FL communication scenarios, are ineffective in addressing these vulnerabilities within sparsified FL. To bridge this gap, we propose FLARE, a novel federated learning framework that integrates sparse index mask inspection and model update sign similarity analysis to detect and mitigate poisoning attacks in sparsified FL. Extensive experiments across multiple datasets and adversarial scenarios demonstrate that FLARE significantly outperforms existing defense strategies, effectively securing sparsified FL against poisoning attacks while maintaining communication efficiency.
Authors: Hendrik Surmann, Jorge de Heuvel, Maren Bennewitz
Abstract: Human drivers exhibit individual preferences regarding driving style. Adapting autonomous vehicles to these preferences is essential for user trust and satisfaction. However, existing end-to-end driving approaches often rely on predefined driving styles or require continuous user feedback for adaptation, limiting their ability to support dynamic, context-dependent preferences. We propose a novel approach using multi-objective reinforcement learning (MORL) with preference-driven optimization for end-to-end autonomous driving that enables runtime adaptation to driving style preferences. Preferences are encoded as continuous weight vectors to modulate behavior along interpretable style objectives$\unicode{x2013}$including efficiency, comfort, speed, and aggressiveness$\unicode{x2013}$without requiring policy retraining. Our single-policy agent integrates vision-based perception in complex mixed-traffic scenarios and is evaluated in diverse urban environments using the CARLA simulator. Experimental results demonstrate that the agent dynamically adapts its driving behavior according to changing preferences while maintaining performance in terms of collision avoidance and route completion.
Authors: Serge Massar
Abstract: We propose a method for training dynamical systems governed by Lagrangian mechanics using Equilibrium Propagation. Our approach extends Equilibrium Propagation - initially developed for energy-based models - to dynamical trajectories by leveraging the principle of action extremization. Training is achieved by gently nudging trajectories toward desired targets and measuring how the variables conjugate to the parameters to be trained respond. This method is particularly suited to systems with periodic boundary conditions or fixed initial and final states, enabling efficient parameter updates without requiring explicit backpropagation through time. In the case of periodic boundary conditions, this approach yields the semiclassical limit of Quantum Equilibrium Propagation. Applications to systems with dissipation are also discussed.
Authors: Zhi Zhong, Akira Takahashi, Shuyang Cui, Keisuke Toyama, Shusuke Takahashi, Yuki Mitsufuji
Abstract: Foley synthesis aims to synthesize high-quality audio that is both semantically and temporally aligned with video frames. Given its broad application in creative industries, the task has gained increasing attention in the research community. To avoid the non-trivial task of training audio generative models from scratch, adapting pretrained audio generative models for video-synchronized foley synthesis presents an attractive direction. ControlNet, a method for adding fine-grained controls to pretrained generative models, has been applied to foley synthesis, but its use has been limited to handcrafted human-readable temporal conditions. In contrast, from-scratch models achieved success by leveraging high-dimensional deep features extracted using pretrained video encoders. We have observed a performance gap between ControlNet-based and from-scratch foley models. To narrow this gap, we propose SpecMaskFoley, a method that steers the pretrained SpecMaskGIT model toward video-synchronized foley synthesis via ControlNet. To unlock the potential of a single ControlNet branch, we resolve the discrepancy between the temporal video features and the time-frequency nature of the pretrained SpecMaskGIT via a frequency-aware temporal feature aligner, eliminating the need for complicated conditioning mechanisms widely used in prior arts. Evaluations on a common foley synthesis benchmark demonstrate that SpecMaskFoley could even outperform strong from-scratch baselines, substantially advancing the development of ControlNet-based foley synthesis models. Demo page: https://zzaudio.github.io/SpecMaskFoley_Demo/
Authors: Parshin Shojaee, Iman Mirzadeh, Keivan Alizadeh, Maxwell Horton, Samy Bengio, Mehrdad Farajtabar
Abstract: Recent generations of language models have introduced Large Reasoning Models (LRMs) that generate detailed thinking processes before providing answers. While these models demonstrate improved performance on reasoning benchmarks, their fundamental capabilities, scaling properties, and limitations remain insufficiently understood. Current evaluations primarily focus on established math and coding benchmarks, emphasizing final answer accuracy. However, this evaluation paradigm often suffers from contamination and does not provide insights into the reasoning traces. In this work, we systematically investigate these gaps with the help of controllable puzzle environments that allow precise manipulation of complexity while maintaining consistent logical structures. This setup enables the analysis of not only final answers but also the internal reasoning traces, offering insights into how LRMs think. Through extensive experiments, we show that LRMs face a complete accuracy collapse beyond certain complexities. Moreover, they exhibit a counterintuitive scaling limit: their reasoning effort increases with problem complexity up to a point, then declines despite having remaining token budget. By comparing LRMs with their standard LLM counterparts under same inference compute, we identify three performance regimes: (1) low-complexity tasks where standard models outperform LRMs, (2) medium-complexity tasks where LRMs demonstrates advantage, and (3) high-complexity tasks where both models face complete collapse. We found that LRMs have limitations in exact computation: they fail to use explicit algorithms and reason inconsistently across scales. We also investigate the reasoning traces in more depth, studying the patterns of explored solutions and analyzing the models' computational behavior, shedding light on their strengths, limitations, and raising questions about their reasoning capabilities.
Authors: Yifan Liu, Yu Fang, Zhouhan Lin
Abstract: Conventional voice conversion modifies voice characteristics from a source speaker to a target speaker, relying on audio input from both sides. However, this process becomes infeasible when clean audio is unavailable, such as in silent videos or noisy environments. In this work, we focus on the task of Silent Face-based Voice Conversion (SFVC), which does voice conversion entirely from visual inputs. i.e., given images of a target speaker and a silent video of a source speaker containing lip motion, SFVC generates speech aligning the identity of the target speaker while preserving the speech content in the source silent video. As this task requires generating intelligible speech and converting identity using only visual cues, it is particularly challenging. To address this, we introduce MuteSwap, a novel framework that employs contrastive learning to align cross-modality identities and minimize mutual information to separate shared visual features. Experimental results show that MuteSwap achieves impressive performance in both speech synthesis and identity conversion, especially under noisy conditions where methods dependent on audio input fail to produce intelligible results, demonstrating both the effectiveness of our training approach and the feasibility of SFVC.
Authors: Matthew A. Chan, Casey J. Pellizzari, Christopher A. Metzler
Abstract: Fourier ptychography (FP) is a powerful light-based synthetic aperture imaging technique that allows one to reconstruct a high-resolution, wide field-of-view image by computationally integrating a diverse collection of low-resolution, far-field measurements. Typically, FP measurement diversity is introduced by changing the angle of the illumination or the position of the camera; either approach results in sampling different portions of the target's spatial frequency content, but both approaches introduce substantial costs and complexity to the acquisition process. In this work, we introduce Inverse Synthetic Aperture Fourier Ptychography, a novel approach to FP that foregoes changing the illumination angle or camera position and instead generates measurement diversity through target motion. Critically, we also introduce a novel learning-based method for estimating k-space coordinates from dual plane intensity measurements, thereby enabling synthetic aperture imaging without knowing the rotation of the target. We experimentally validate our method in simulation and on a tabletop optical system.
Authors: Sarthak Choudhary, Divyam Anshumaan, Nils Palumbo, Somesh Jha
Abstract: LLM-integrated applications and agents are vulnerable to prompt injection attacks, in which adversaries embed malicious instructions within seemingly benign user inputs to manipulate the LLM's intended behavior. Recent defenses based on $\textit{known-answer detection}$ (KAD) have achieved near-perfect performance by using an LLM to classify inputs as clean or contaminated. In this work, we formally characterize the KAD framework and uncover a structural vulnerability in its design that invalidates its core security premise. We design a methodical adaptive attack, $\textit{DataFlip}$, to exploit this fundamental weakness. It consistently evades KAD defenses with detection rates as low as $1.5\%$ while reliably inducing malicious behavior with success rates of up to $88\%$, without needing white-box access to the LLM or any optimization procedures.
Authors: Ruihan Yang, Qinxi Yu, Yecheng Wu, Rui Yan, Borui Li, An-Chieh Cheng, Xueyan Zou, Yunhao Fang, Xuxin Cheng, Ri-Zhao Qiu, Hongxu Yin, Sifei Liu, Song Han, Yao Lu, Xiaolong Wang
Abstract: Real robot data collection for imitation learning has led to significant advancements in robotic manipulation. However, the requirement for robot hardware in the process fundamentally constrains the scale of the data. In this paper, we explore training Vision-Language-Action (VLA) models using egocentric human videos. The benefit of using human videos is not only for their scale but more importantly for the richness of scenes and tasks. With a VLA trained on human video that predicts human wrist and hand actions, we can perform Inverse Kinematics and retargeting to convert the human actions to robot actions. We fine-tune the model using a few robot manipulation demonstrations to obtain the robot policy, namely EgoVLA. We propose a simulation benchmark called Ego Humanoid Manipulation Benchmark, where we design diverse bimanual manipulation tasks with demonstrations. We fine-tune and evaluate EgoVLA with Ego Humanoid Manipulation Benchmark and show significant improvements over baselines and ablate the importance of human data. Videos can be found on our website: https://rchalyang.github.io/EgoVLA
Authors: Keishi Sando, Hideitsu Hino
Abstract: Graph representation matrices are essential tools in graph data analysis. Recently, Hermitian adjacency matrices have been proposed to investigate directed graph structures. Previous studies have demonstrated that these matrices can extract valuable information for clustering. In this paper, we propose the complex non-backtracking matrix that integrates the properties of the Hermitian adjacency matrix and the non-backtracking matrix. The proposed matrix has similar properties with the non-backtracking matrix of undirected graphs. We reveal relationships between the complex non-backtracking matrix and the Hermitian adjacency matrix. Also, we provide intriguing insights that this matrix representation holds cluster information, particularly for sparse directed graphs.
Authors: Ammar Ahmed, Ali Shariq Imran, Zenun Kastrati, Sher Muhammad Daudpota
Abstract: Wrist pathologies are frequently observed, particularly among children who constitute the majority of fracture cases. However, diagnosing these conditions is time-consuming and requires specialized expertise. Computer vision presents a promising avenue, contingent upon the availability of extensive datasets, a notable challenge in medical imaging. Therefore, reliance solely on one modality, such as images, proves inadequate, especially in an era of diverse and plentiful data types. In this study, we employ a multifaceted approach to address the challenge of recognizing wrist pathologies using an extremely limited dataset. Initially, we approach the problem as a fine-grained recognition task, aiming to identify subtle X-ray pathologies that conventional CNNs overlook. Secondly, we enhance network performance by fusing patient metadata with X-ray images. Thirdly, rather than pre-training on a coarse-grained dataset like ImageNet, we utilize weights trained on a fine-grained dataset. While metadata integration has been used in other medical domains, this is a novel application for wrist pathologies. Our results show that a fine-grained strategy and metadata integration improve diagnostic accuracy by 2% with a limited dataset and by over 10% with a larger fracture-focused dataset.