id
stringlengths
9
16
title
stringlengths
4
278
categories
stringlengths
5
104
abstract
stringlengths
6
4.09k
2502.09303
Towards Seamless Hierarchical Federated Learning under Intermittent Client Participation: A Stagewise Decision-Making Methodology
cs.LG cs.DC
Federated Learning (FL) offers a pioneering distributed learning paradigm that enables devices/clients to build a shared global model. This global model is obtained through frequent model transmissions between clients and a central server, which may cause high latency, energy consumption, and congestion over backhaul links. To overcome these drawbacks, Hierarchical Federated Learning (HFL) has emerged, which organizes clients into multiple clusters and utilizes edge nodes (e.g., edge servers) for intermediate model aggregations between clients and the central server. Current research on HFL mainly focus on enhancing model accuracy, latency, and energy consumption in scenarios with a stable/fixed set of clients. However, addressing the dynamic availability of clients -- a critical aspect of real-world scenarios -- remains underexplored. This study delves into optimizing client selection and client-to-edge associations in HFL under intermittent client participation so as to minimize overall system costs (i.e., delay and energy), while achieving fast model convergence. We unveil that achieving this goal involves solving a complex NP-hard problem. To tackle this, we propose a stagewise methodology that splits the solution into two stages, referred to as Plan A and Plan B. Plan A focuses on identifying long-term clients with high chance of participation in subsequent model training rounds. Plan B serves as a backup, selecting alternative clients when long-term clients are unavailable during model training rounds. This stagewise methodology offers a fresh perspective on client selection that can enhance both HFL and conventional FL via enabling low-overhead decision-making processes. Through evaluations on MNIST and CIFAR-10 datasets, we show that our methodology outperforms existing benchmarks in terms of model accuracy and system costs.
2502.09304
KET-RAG: A Cost-Efficient Multi-Granular Indexing Framework for Graph-RAG
cs.IR
Graph-RAG constructs a knowledge graph from text chunks to improve retrieval in Large Language Model (LLM)-based question answering. It is particularly useful in domains such as biomedicine, law, and political science, where retrieval often requires multi-hop reasoning over proprietary documents. Some existing Graph-RAG systems construct KNN graphs based on text chunk relevance, but this coarse-grained approach fails to capture entity relationships within texts, leading to sub-par retrieval and generation quality. To address this, recent solutions leverage LLMs to extract entities and relationships from text chunks, constructing triplet-based knowledge graphs. However, this approach incurs significant indexing costs, especially for large document collections. To ensure a good result accuracy while reducing the indexing cost, we propose KET-RAG, a multi-granular indexing framework. KET-RAG first identifies a small set of key text chunks and leverages an LLM to construct a knowledge graph skeleton. It then builds a text-keyword bipartite graph from all text chunks, serving as a lightweight alternative to a full knowledge graph. During retrieval, KET-RAG searches both structures: it follows the local search strategy of existing Graph-RAG systems on the skeleton while mimicking this search on the bipartite graph to improve retrieval quality. We evaluate eight solutions on two real-world datasets, demonstrating that KET-RAG outperforms all competitors in indexing cost, retrieval effectiveness, and generation quality. Notably, it achieves comparable or superior retrieval quality to Microsoft's Graph-RAG while reducing indexing costs by over an order of magnitude. Additionally, it improves the generation quality by up to 32.4% while lowering indexing costs by around 20%.
2502.09305
Predicting Drive Test Results in Mobile Networks Using Optimization Techniques
cs.NI cs.AI cs.SE
Mobile network operators constantly optimize their networks to ensure superior service quality and coverage. This optimization is crucial for maintaining an optimal user experience and requires extensive data collection and analysis. One of the primary methods for gathering this data is through drive tests, where technical teams use specialized equipment to collect signal information across various regions. However, drive tests are both costly and time-consuming, and they face challenges such as traffic conditions, environmental factors, and limited access to certain areas. These constraints make it difficult to replicate drive tests under similar conditions. In this study, we propose a method that enables operators to predict received signal strength at specific locations using data from other drive test points. By reducing the need for widespread drive tests, this approach allows operators to save time and resources while still obtaining the necessary data to optimize their networks and mitigate the challenges associated with traditional drive tests.
2502.09306
Non-asymptotic Analysis of Diffusion Annealed Langevin Monte Carlo for Generative Modelling
stat.ML cs.LG math.PR stat.CO
We investigate the theoretical properties of general diffusion (interpolation) paths and their Langevin Monte Carlo implementation, referred to as diffusion annealed Langevin Monte Carlo (DALMC), under weak conditions on the data distribution. Specifically, we analyse and provide non-asymptotic error bounds for the annealed Langevin dynamics where the path of distributions is defined as Gaussian convolutions of the data distribution as in diffusion models. We then extend our results to recently proposed heavy-tailed (Student's t) diffusion paths, demonstrating their theoretical properties for heavy-tailed data distributions for the first time. Our analysis provides theoretical guarantees for a class of score-based generative models that interpolate between a simple distribution (Gaussian or Student's t) and the data distribution in finite time. This approach offers a broader perspective compared to standard score-based diffusion approaches, which are typically based on a forward Ornstein-Uhlenbeck (OU) noising process.
2502.09307
When the LM misunderstood the human chuckled: Analyzing garden path effects in humans and language models
cs.CL cs.AI
Modern Large Language Models (LLMs) have shown human-like abilities in many language tasks, sparking interest in comparing LLMs' and humans' language processing. In this paper, we conduct a detailed comparison of the two on a sentence comprehension task using garden-path constructions, which are notoriously challenging for humans. Based on psycholinguistic research, we formulate hypotheses on why garden-path sentences are hard, and test these hypotheses on human participants and a large suite of LLMs using comprehension questions. Our findings reveal that both LLMs and humans struggle with specific syntactic complexities, with some models showing high correlation with human comprehension. To complement our findings, we test LLM comprehension of garden-path constructions with paraphrasing and text-to-image generation tasks, and find that the results mirror the sentence comprehension question results, further validating our findings on LLM understanding of these constructions.
2502.09309
Frequency Domain Stability and Convergence Analysis for General Reset Control Systems Architecture
eess.SY cs.SY
A key factor that generates significant interest in reset control systems, especially within industrial contexts, is their potential to be designed using a frequency-domain loop-shaping procedure. On the other hand, formulating and assessing stability analysis for these nonlinear elements often depends on access to parametric models and numerically solving linear matrix inequalities. These specific factors could present challenges to the successful implementation of reset control within industrial settings. Moreover, one of the most effective structures for implementing reset elements is to use them in parallel with a linear element. Therefore, this article presents the development of the frequency domain-based $H_\beta$ stability method from a series to a more general structure of reset control systems. Additionally, it investigates the behavior of different reset elements in terms of the feasibility of stability in the presence of time delay. To illustrate the research findings, two examples are provided, including one from an industrial application.
2502.09310
Global Stabilization of Chemostats with Nonzero Mortality and Substrate Dynamics
math.OC cs.SY eess.SY q-bio.PE
In "chemostat"-type population models that incorporate substrate (nutrient) dynamics, the dependence of the birth (or growth) rate on the substrate concentration introduces nonlinear coupling that creates a challenge for stabilization that is global, namely, for all positive concentrations of the biomass and nutrients. This challenge for global stabilization has been overcome in the literature using relatively simple feedback when natural mortality of the biomass is absent. However, under natural mortality, it takes fortified, more complex feedback, outside of the existing nonlinear control design toolbox, to avoid biomass extinction from nutrient-depleted initial conditions. Such fortified feedback, the associated control Laypunov function design, and Lyapunov analysis of global stability are provided in this paper. We achieve global stabilization for two different chemostat models: (i) a lumped model, with two state variables, and (ii) a three-state model derived from an age-structured infinite-dimensional model. The proposed feedback stabilizers are explicit, applicable to both the lumped and the age-structured models, and coincide with simple feedback laws proposed in the literature when the mortality rate is zero. Global stabilization means subject to constraints: all positive biomass and nutrient concentrations are within the region of attraction of the desired equilibrium, and, additionally, this is achieved with a dilution input that is guaranteed to remain positive. For the lumped case with Haldane kinetics, we show that the reproduction rate dominating the mortality (excluding the reproduction and mortality being in balance) is not only sufficient but also necessary for global stabilization. The obtained results are illustrated with simple examples.
2502.09311
Mitigating the Impact of Prominent Position Shift in Drone-based RGBT Object Detection
cs.CV
Drone-based RGBT object detection plays a crucial role in many around-the-clock applications. However, real-world drone-viewed RGBT data suffers from the prominent position shift problem, i.e., the position of a tiny object differs greatly in different modalities. For instance, a slight deviation of a tiny object in the thermal modality will induce it to drift from the main body of itself in the RGB modality. Considering RGBT data are usually labeled on one modality (reference), this will cause the unlabeled modality (sensed) to lack accurate supervision signals and prevent the detector from learning a good representation. Moreover, the mismatch of the corresponding feature point between the modalities will make the fused features confusing for the detection head. In this paper, we propose to cast the cross-modality box shift issue as the label noise problem and address it on the fly via a novel Mean Teacher-based Cross-modality Box Correction head ensemble (CBC). In this way, the network can learn more informative representations for both modalities. Furthermore, to alleviate the feature map mismatch problem in RGBT fusion, we devise a Shifted Window-Based Cascaded Alignment (SWCA) module. SWCA mines long-range dependencies between the spatially unaligned features inside shifted windows and cascaded aligns the sensed features with the reference ones. Extensive experiments on two drone-based RGBT object detection datasets demonstrate that the correction results are both visually and quantitatively favorable, thereby improving the detection performance. In particular, our CBC module boosts the precision of the sensed modality ground truth by 25.52 aSim points. Overall, the proposed detector achieves an mAP_50 of 43.55 points on RGBTDronePerson and surpasses a state-of-the-art method by 8.6 mAP50 on a shift subset of DroneVehicle dataset. The code and data will be made publicly available.
2502.09316
A Judge-free LLM Open-ended Generation Benchmark Based on the Distributional Hypothesis
cs.CL
Evaluating the open-ended text generation of large language models (LLMs) is challenging because of the lack of a clear ground truth and the high cost of human or LLM-based assessments. We propose a novel benchmark that evaluates LLMs using n-gram statistics and rules, without relying on human judgement or LLM-as-a-judge approaches. Using 50 question and reference answer sets, we introduce three new metrics based on n-grams and rules: Fluency, Truthfulness, and Helpfulness. Our benchmark strongly correlates with GPT-4o-based evaluations while requiring significantly fewer computational resources, demonstrating its effectiveness as a scalable alternative for assessing LLMs' open-ended generation capabilities.
2502.09318
SigGate: Enhancing Recurrent Neural Networks with Signature-Based Gating Mechanisms
cs.LG
In this paper, we propose a novel approach that enhances recurrent neural networks (RNNs) by incorporating path signatures into their gating mechanisms. Our method modifies both Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) architectures by replacing their forget and reset gates, respectively, with learnable path signatures. These signatures, which capture the geometric features of the entire path history, provide a richer context for controlling information flow through the network's memory. This modification allows the networks to make memory decisions based on the full historical context rather than just the current input and state. Through experimental studies, we demonstrate that our Signature-LSTM (SigLSTM) and Signature-GRU (SigGRU) models outperform their traditional counterparts across various sequential learning tasks. By leveraging path signatures in recurrent architectures, this method offers new opportunities to enhance performance in time series analysis and forecasting applications.
2502.09319
Bridging Jensen Gap for Max-Min Group Fairness Optimization in Recommendation
cs.IR cs.LG
Group max-min fairness (MMF) is commonly used in fairness-aware recommender systems (RS) as an optimization objective, as it aims to protect marginalized item groups and ensures a fair competition platform. However, our theoretical analysis indicates that integrating MMF constraint violates the assumption of sample independence during optimization, causing the loss function to deviate from linear additivity. Such nonlinearity property introduces the Jensen gap between the model's convergence point and the optimal point if mini-batch sampling is applied. Both theoretical and empirical studies show that as the mini-batch size decreases and the group size increases, the Jensen gap will widen accordingly. Some methods using heuristic re-weighting or debiasing strategies have the potential to bridge the Jensen gap. However, they either lack theoretical guarantees or suffer from heavy computational costs. To overcome these limitations, we first theoretically demonstrate that the MMF-constrained objective can be essentially reformulated as a group-weighted optimization objective. Then we present an efficient and effective algorithm named FairDual, which utilizes a dual optimization technique to minimize the Jensen gap. Our theoretical analysis demonstrates that FairDual can achieve a sub-linear convergence rate to the globally optimal solution and the Jensen gap can be well bounded under a mini-batch sampling strategy with random shuffle. Extensive experiments conducted using six large-scale RS backbone models on three publicly available datasets demonstrate that FairDual outperforms all baselines in terms of both accuracy and fairness. Our data and codes are shared at https://github.com/XuChen0427/FairDual.
2502.09324
Depth-Bounds for Neural Networks via the Braid Arrangement
cs.LG cs.DM cs.NE math.CO
We contribute towards resolving the open question of how many hidden layers are required in ReLU networks for exactly representing all continuous and piecewise linear functions on $\mathbb{R}^d$. While the question has been resolved in special cases, the best known lower bound in general is still 2. We focus on neural networks that are compatible with certain polyhedral complexes, more precisely with the braid fan. For such neural networks, we prove a non-constant lower bound of $\Omega(\log\log d)$ hidden layers required to exactly represent the maximum of $d$ numbers. Additionally, under our assumption, we provide a combinatorial proof that 3 hidden layers are necessary to compute the maximum of 5 numbers; this had only been verified with an excessive computation so far. Finally, we show that a natural generalization of the best known upper bound to maxout networks is not tight, by demonstrating that a rank-3 maxout layer followed by a rank-2 maxout layer is sufficient to represent the maximum of 7 numbers.
2502.09325
A Benchmark for Crime Surveillance Video Analysis with Large Models
cs.CV
Anomaly analysis in surveillance videos is a crucial topic in computer vision. In recent years, multimodal large language models (MLLMs) have outperformed task-specific models in various domains. Although MLLMs are particularly versatile, their abilities to understand anomalous concepts and details are insufficiently studied because of the outdated benchmarks of this field not providing MLLM-style QAs and efficient algorithms to assess the model's open-ended text responses. To fill this gap, we propose a benchmark for crime surveillance video analysis with large models denoted as UCVL, including 1,829 videos and reorganized annotations from the UCF-Crime and UCF-Crime Annotation datasets. We design six types of questions and generate diverse QA pairs. Then we develop detailed instructions and use OpenAI's GPT-4o for accurate assessment. We benchmark eight prevailing MLLMs ranging from 0.5B to 40B parameters, and the results demonstrate the reliability of this bench. Moreover, we finetune LLaVA-OneVision on UCVL's training set. The improvement validates our data's high quality for video anomaly analysis.
2502.09329
Bayesian Optimization for Simultaneous Selection of Machine Learning Algorithms and Hyperparameters on Shared Latent Space
cs.LG
Selecting the optimal combination of a machine learning (ML) algorithm and its hyper-parameters is crucial for the development of high-performance ML systems. However, since the combination of ML algorithms and hyper-parameters is enormous, the exhaustive validation requires a significant amount of time. Many existing studies use Bayesian optimization (BO) for accelerating the search. On the other hand, a significant difficulty is that, in general, there exists a different hyper-parameter space for each one of candidate ML algorithms. BO-based approaches typically build a surrogate model independently for each hyper-parameter space, by which sufficient observations are required for all candidate ML algorithms. In this study, our proposed method embeds different hyper-parameter spaces into a shared latent space, in which a surrogate multi-task model for BO is estimated. This approach can share information of observations from different ML algorithms by which efficient optimization is expected with a smaller number of total observations. We further propose the pre-training of the latent space embedding with an adversarial regularization, and a ranking model for selecting an effective pre-trained embedding for a given target dataset. Our empirical study demonstrates effectiveness of the proposed method through datasets from OpenML.
2502.09331
Beyond English: The Impact of Prompt Translation Strategies across Languages and Tasks in Multilingual LLMs
cs.CL
Despite advances in the multilingual capabilities of Large Language Models (LLMs) across diverse tasks, English remains the dominant language for LLM research and development. So, when working with a different language, this has led to the widespread practice of pre-translation, i.e., translating the task prompt into English before inference. Selective pre-translation, a more surgical approach, focuses on translating specific prompt components. However, its current use is sporagic and lacks a systematic research foundation. Consequently, the optimal pre-translation strategy for various multilingual settings and tasks remains unclear. In this work, we aim to uncover the optimal setup for pre-translation by systematically assessing its use. Specifically, we view the prompt as a modular entity, composed of four functional parts: instruction, context, examples, and output, either of which could be translated or not. We evaluate pre-translation strategies across 35 languages covering both low and high-resource languages, on various tasks including Question Answering (QA), Natural Language Inference (NLI), Named Entity Recognition (NER), and Abstractive Summarization. Our experiments show the impact of factors as similarity to English, translation quality and the size of pre-trained data, on the model performance with pre-translation. We suggest practical guidelines for choosing optimal strategies in various multilingual settings.
2502.09332
Full Swap Regret and Discretized Calibration
cs.LG cs.GT
We study the problem of minimizing swap regret in structured normal-form games. Players have a very large (potentially infinite) number of pure actions, but each action has an embedding into $d$-dimensional space and payoffs are given by bilinear functions of these embeddings. We provide an efficient learning algorithm for this setting that incurs at most $\tilde{O}(T^{(d+1)/(d+3)})$ swap regret after $T$ rounds. To achieve this, we introduce a new online learning problem we call \emph{full swap regret minimization}. In this problem, a learner repeatedly takes a (randomized) action in a bounded convex $d$-dimensional action set $\mathcal{K}$ and then receives a loss from the adversary, with the goal of minimizing their regret with respect to the \emph{worst-case} swap function mapping $\mathcal{K}$ to $\mathcal{K}$. For varied assumptions about the convexity and smoothness of the loss functions, we design algorithms with full swap regret bounds ranging from $O(T^{d/(d+2)})$ to $O(T^{(d+1)/(d+2)})$. Finally, we apply these tools to the problem of online forecasting to minimize calibration error, showing that several notions of calibration can be viewed as specific instances of full swap regret. In particular, we design efficient algorithms for online forecasting that guarantee at most $O(T^{1/3})$ $\ell_2$-calibration error and $O(\max(\sqrt{\epsilon T}, T^{1/3}))$ \emph{discretized-calibration} error (when the forecaster is restricted to predicting multiples of $\epsilon$).
2502.09335
Graph Diffusion Network for Drug-Gene Prediction
cs.LG cs.AI
Predicting drug-gene associations is crucial for drug development and disease treatment. While graph neural networks (GNN) have shown effectiveness in this task, they face challenges with data sparsity and efficient contrastive learning implementation. We introduce a graph diffusion network for drug-gene prediction (GDNDGP), a framework that addresses these limitations through two key innovations. First, it employs meta-path-based homogeneous graph learning to capture drug-drug and gene-gene relationships, ensuring similar entities share embedding spaces. Second, it incorporates a parallel diffusion network that generates hard negative samples during training, eliminating the need for exhaustive negative sample retrieval. Our model achieves superior performance on the DGIdb 4.0 dataset and demonstrates strong generalization capability on tripartite drug-gene-disease networks. Results show significant improvements over existing methods in drug-gene prediction tasks, particularly in handling complex heterogeneous relationships. The source code is publicly available at https://github.com/csjywu1/GDNDGP.
2502.09340
This looks like what? Challenges and Future Research Directions for Part-Prototype Models
cs.LG
The growing interest in eXplainable Artificial Intelligence (XAI) has prompted research into models with built-in interpretability, the most prominent of which are part-prototype models. Part-Prototype Models (PPMs) make decisions by comparing an input image to a set of learned prototypes, providing human-understandable explanations in the form of ``this looks like that''. Despite their inherent interpretability, PPMS are not yet considered a valuable alternative to post-hoc models. In this survey, we investigate the reasons for this and provide directions for future research. We analyze papers from 2019 to 2024, and derive a taxonomy of the challenges that current PPMS face. Our analysis shows that the open challenges are quite diverse. The main concern is the quality and quantity of prototypes. Other concerns are the lack of generalization to a variety of tasks and contexts, and general methodological issues, including non-standardized evaluation. We provide ideas for future research in five broad directions: improving predictive performance, developing novel architectures grounded in theory, establishing frameworks for human-AI collaboration, aligning models with humans, and establishing metrics and benchmarks for evaluation. We hope that this survey will stimulate research and promote intrinsically interpretable models for application domains. Our list of surveyed papers is available at https://github.com/aix-group/ppm-survey.
2502.09341
Neural Spatiotemporal Point Processes: Trends and Challenges
cs.LG cs.AI
Spatiotemporal point processes (STPPs) are probabilistic models for events occurring in continuous space and time. Real-world event data often exhibit intricate dependencies and heterogeneous dynamics. By incorporating modern deep learning techniques, STPPs can model these complexities more effectively than traditional approaches. Consequently, the fusion of neural methods with STPPs has become an active and rapidly evolving research area. In this review, we categorize existing approaches, unify key design choices, and explain the challenges of working with this data modality. We further highlight emerging trends and diverse application domains. Finally, we identify open challenges and gaps in the literature.
2502.09344
Revisiting Topological Interference Management: A Learning-to-Code on Graphs Perspective
cs.IT math.IT
The advance of topological interference management (TIM) has been one of the driving forces of recent developments in network information theory. However, state-of-the-art coding schemes for TIM are usually handcrafted for specific families of network topologies, relying critically on experts' domain knowledge and sophisticated treatments. The lack of systematic and automatic generation of solutions inevitably restricts their potential wider applications to wireless communication systems, due to the limited generalizability of coding schemes to wider network configurations. To address such an issue, this work makes the first attempt to advocate revisiting topological interference alignment (IA) from a novel learning-to-code perspective. Specifically, we recast the one-to-one and subspace IA conditions as vector assignment policies and propose a unifying learning-to-code on graphs (LCG) framework by leveraging graph neural networks (GNNs) for capturing topological structures and reinforcement learning (RL) for decision-making of IA beamforming vector assignment. Interestingly, the proposed LCG framework is capable of recovering known one-to-one scalar/vector IA solutions for a significantly wider range of network topologies, and more remarkably of discovering new subspace IA coding schemes for multiple-antenna cases that are challenging to be handcrafted. The extensive experiments demonstrate that the LCG framework is an effective way to automatically produce systematic coding solutions to the TIM instances with arbitrary network topologies, and at the same time, the underlying learning algorithm is efficient with respect to online inference time and possesses excellent generalizability and transferability for practical deployment.
2502.09346
Machine learning for modelling unstructured grid data in computational physics: a review
cs.LG cs.CE physics.data-an physics.flu-dyn
Unstructured grid data are essential for modelling complex geometries and dynamics in computational physics. Yet, their inherent irregularity presents significant challenges for conventional machine learning (ML) techniques. This paper provides a comprehensive review of advanced ML methodologies designed to handle unstructured grid data in high-dimensional dynamical systems. Key approaches discussed include graph neural networks, transformer models with spatial attention mechanisms, interpolation-integrated ML methods, and meshless techniques such as physics-informed neural networks. These methodologies have proven effective across diverse fields, including fluid dynamics and environmental simulations. This review is intended as a guidebook for computational scientists seeking to apply ML approaches to unstructured grid data in their domains, as well as for ML researchers looking to address challenges in computational physics. It places special focus on how ML methods can overcome the inherent limitations of traditional numerical techniques and, conversely, how insights from computational physics can inform ML development. To support benchmarking, this review also provides a summary of open-access datasets of unstructured grid data in computational physics. Finally, emerging directions such as generative models with unstructured data, reinforcement learning for mesh generation, and hybrid physics-data-driven paradigms are discussed to inspire future advancements in this evolving field.
2502.09352
Wasserstein distributional adversarial training for deep neural networks
cs.LG cs.CV math.OC
Design of adversarial attacks for deep neural networks, as well as methods of adversarial training against them, are subject of intense research. In this paper, we propose methods to train against distributional attack threats, extending the TRADES method used for pointwise attacks. Our approach leverages recent contributions and relies on sensitivity analysis for Wasserstein distributionally robust optimization problems. We introduce an efficient fine-tuning method which can be deployed on a previously trained model. We test our methods on a range of pre-trained models on RobustBench. These experimental results demonstrate the additional training enhances Wasserstein distributional robustness, while maintaining original levels of pointwise robustness, even for already very successful networks. The improvements are less marked for models pre-trained using huge synthetic datasets of 20-100M images. However, remarkably, sometimes our methods are still able to improve their performance even when trained using only the original training dataset (50k images).
2502.09355
Simultaneous solution of incompressible Navier-Stokes flows on multiple surfaces
cs.CE
A mechanical model and finite element method for the simultaneous solution of Stokes and incompressible Navier-Stokes flows on multiple curved surfaces over a bulk domain are proposed. The two-dimensional surfaces are defined implicitly by all level sets of a scalar function, bounded by the three-dimensional bulk domain. This bulk domain is discretized with hexahedral finite elements which do not necessarily conform with the level sets but with the boundary. The resulting numerical method is a hybrid between conforming and non-conforming finite element methods. Taylor-Hood elements or equal-order element pairs for velocity and pressure, together with stabilization techniques, are applied to fulfil the inf-sup conditions resulting from the mixed-type formulation of the governing equations. Numerical studies confirm good agreement with independently obtained solutions on selected, individual surfaces. Furthermore, higher-order convergence rates are obtained for sufficiently smooth solutions.
2502.09356
Galileo: Learning Global and Local Features in Pretrained Remote Sensing Models
cs.CV
From crop mapping to flood detection, machine learning in remote sensing has a wide range of societally beneficial applications. The commonalities between remote sensing data in these applications present an opportunity for pretrained machine learning models tailored to remote sensing to reduce the labeled data and effort required to solve individual tasks. However, such models must be: (i) flexible enough to ingest input data of varying sensor modalities and shapes (i.e., of varying spatial and temporal dimensions), and (ii) able to model Earth surface phenomena of varying scales and types. To solve this gap, we present Galileo, a family of pretrained remote sensing models designed to flexibly process multimodal remote sensing data. We also introduce a novel and highly effective self-supervised learning approach to learn both large- and small-scale features, a challenge not addressed by previous models. Our Galileo models obtain state-of-the-art results across diverse remote sensing tasks.
2502.09363
The Accuracy Cost of Weakness: A Theoretical Analysis of Fixed-Segment Weak Labeling for Events in Time
cs.LG
Accurate labels are critical for deriving robust machine learning models. Labels are used to train supervised learning models and to evaluate most machine learning paradigms. In this paper, we model the accuracy and cost of a common weak labeling process where annotators assign presence or absence labels to fixed-length data segments for a given event class. The annotator labels a segment as "present" if it sufficiently covers an event from that class, e.g., a birdsong sound event in audio data. We analyze how the segment length affects the label accuracy and the required number of annotations, and compare this fixed-length labeling approach with an oracle method that uses the true event activations to construct the segments. Furthermore, we quantify the gap between these methods and verify that in most realistic scenarios the oracle method is better than the fixed-length labeling method in both accuracy and cost. Our findings provide a theoretical justification for adaptive weak labeling strategies that mimic the oracle process, and a foundation for optimizing weak labeling processes in sequence labeling tasks.
2502.09365
Simple Path Structural Encoding for Graph Transformers
cs.LG cs.AI
Graph transformers extend global self-attention to graph-structured data, achieving notable success in graph learning. Recently, random walk structural encoding (RWSE) has been found to further enhance their predictive power by encoding both structural and positional information into the edge representation. However, RWSE cannot always distinguish between edges that belong to different local graph patterns, which reduces its ability to capture the full structural complexity of graphs. This work introduces Simple Path Structural Encoding (SPSE), a novel method that utilizes simple path counts for edge encoding. We show theoretically and experimentally that SPSE overcomes the limitations of RWSE, providing a richer representation of graph structures, particularly for capturing local cyclic patterns. To make SPSE computationally tractable, we propose an efficient approximate algorithm for simple path counting. SPSE demonstrates significant performance improvements over RWSE on various benchmarks, including molecular and long-range graph datasets, achieving statistically significant gains in discriminative tasks. These results pose SPSE as a powerful edge encoding alternative for enhancing the expressivity of graph transformers.
2502.09368
Optimal Microcontroller Usage in Reconfigurable Intelligent Surface: Batteryless IoT Systems Case Study
cs.IT math.IT
To enhance wireless communication in IoT systems using reconfigurable intelligent surfaces (RISs), efficient control of programmable passive and active elements is essential. However, increasing RIS elements requires more microcontrollers, raising complexity and cost. This paper proposes a modular approach ("Module"), where each microcontroller controls a module of optimal active or passive elements. The module size is determined using a non-linear energy harvesting model, where a batteryless IoT (b-IoT) sensor harvests energy from base station (BS) RF signals. We optimize the number of modules (microcontrollers) to minimize energy consumption while satisfying energy harvesting and information causality constraints. Simulations show that RIS module-assisted energy harvesting improves IoT system performance by ~100% compared to models without RIS panels.
2502.09369
Language Agents as Digital Representatives in Collective Decision-Making
cs.LG cs.AI cs.CL cs.CY
Consider the process of collective decision-making, in which a group of individuals interactively select a preferred outcome from among a universe of alternatives. In this context, "representation" is the activity of making an individual's preferences present in the process via participation by a proxy agent -- i.e. their "representative". To this end, learned models of human behavior have the potential to fill this role, with practical implications for multi-agent scenario studies and mechanism design. In this work, we investigate the possibility of training \textit{language agents} to behave in the capacity of representatives of human agents, appropriately expressing the preferences of those individuals whom they stand for. First, we formalize the setting of \textit{collective decision-making} -- as the episodic process of interaction between a group of agents and a decision mechanism. On this basis, we then formalize the problem of \textit{digital representation} -- as the simulation of an agent's behavior to yield equivalent outcomes from the mechanism. Finally, we conduct an empirical case study in the setting of \textit{consensus-finding} among diverse humans, and demonstrate the feasibility of fine-tuning large language models to act as digital representatives.
2502.09374
Mitigating multiple single-event upsets during deep neural network inference using fault-aware training
cs.LG
Deep neural networks (DNNs) are increasingly used in safety-critical applications. Reliable fault analysis and mitigation are essential to ensure their functionality in harsh environments that contain high radiation levels. This study analyses the impact of multiple single-bit single-event upsets in DNNs by performing fault injection at the level of a DNN model. Additionally, a fault aware training (FAT) methodology is proposed that improves the DNNs' robustness to faults without any modification to the hardware. Experimental results show that the FAT methodology improves the tolerance to faults up to a factor 3.
2502.09375
FARM: Frequency-Aware Model for Cross-Domain Live-Streaming Recommendation
cs.IR
Live-streaming services have attracted widespread popularity due to their real-time interactivity and entertainment value. Users can engage with live-streaming authors by participating in live chats, posting likes, or sending virtual gifts to convey their preferences and support. However, the live-streaming services faces serious data-sparsity problem, which can be attributed to the following two points: (1) User's valuable behaviors are usually sparse, e.g., like, comment and gift, which are easily overlooked by the model, making it difficult to describe user's personalized preference. (2) The main exposure content on our platform is short-video, which is 9 times higher than the exposed live-streaming, leading to the inability of live-streaming content to fully model user preference. To this end, we propose a Frequency-Aware Model for Cross-Domain Live-Streaming Recommendation, termed as FARM. Specifically, we first present the intra-domain frequency aware module to enable our model to perceive user's sparse yet valuable behaviors, i.e., high-frequency information, supported by the Discrete Fourier Transform (DFT). To transfer user preference across the short-video and live-streaming domains, we propose a novel preference align before fuse strategy, which consists of two parts: the cross-domain preference align module to align user preference in both domains with contrastive learning, and the cross-domain preference fuse module to further fuse user preference in both domains using a serious of tailor-designed attention mechanisms. Extensive offline experiments and online A/B testing on Kuaishou live-streaming services demonstrate the effectiveness and superiority of FARM. Our FARM has been deployed in online live-streaming services and currently serves hundreds of millions of users on Kuaishou.
2502.09376
LoRA Training Provably Converges to a Low-Rank Global Minimum or It Fails Loudly (But it Probably Won't Fail)
cs.LG
Low-rank adaptation (LoRA) has become a standard approach for fine-tuning large foundation models. However, our theoretical understanding of LoRA remains limited as prior analyses of LoRA's training dynamics either rely on linearization arguments or consider highly simplified setups. In this work, we analyze the LoRA loss landscape without such restrictive assumptions. We define two regimes: a ``special regime'', which includes idealized setups where linearization arguments hold, and a ``generic regime'' representing more realistic setups where linearization arguments do not hold. In the generic regime, we show that LoRA training converges to a global minimizer with low rank and small magnitude, or a qualitatively distinct solution with high rank and large magnitude. Finally, we argue that the zero-initialization and weight decay in LoRA training induce an implicit bias toward the low-rank, small-magnitude region of the parameter space -- where global minima lie -- thus shedding light on why LoRA training usually succeeds in finding global minima.
2502.09378
A Deep Inverse-Mapping Model for a Flapping Robotic Wing
cs.AI cs.RO
In systems control, the dynamics of a system are governed by modulating its inputs to achieve a desired outcome. For example, to control the thrust of a quad-copter propeller the controller modulates its rotation rate, relying on a straightforward mapping between the input rotation rate and the resulting thrust. This mapping can be inverted to determine the rotation rate needed to generate a desired thrust. However, in complex systems, such as flapping-wing robots where intricate fluid motions are involved, mapping inputs (wing kinematics) to outcomes (aerodynamic forces) is nontrivial and inverting this mapping for real-time control is computationally impractical. Here, we report a machine-learning solution for the inverse mapping of a flapping-wing system based on data from an experimental system we have developed. Our model learns the input wing motion required to generate a desired aerodynamic force outcome. We used a sequence-to-sequence model tailored for time-series data and augmented it with a novel adaptive-spectrum layer that implements representation learning in the frequency domain. To train our model, we developed a flapping wing system that simultaneously measures the wing's aerodynamic force and its 3D motion using high-speed cameras. We demonstrate the performance of our system on an additional open-source dataset of a flapping wing in a different flow regime. Results show superior performance compared with more complex state-of-the-art transformer-based models, with 11% improvement on the test datasets median loss. Moreover, our model shows superior inference time, making it practical for onboard robotic control. Our open-source data and framework may improve modeling and real-time control of systems governed by complex dynamics, from biomimetic robots to biomedical devices.
2502.09379
TRIFFID: Autonomous Robotic Aid For Increasing First Responders Efficiency
cs.RO cs.AI
The increasing complexity of natural disaster incidents demands innovative technological solutions to support first responders in their efforts. This paper introduces the TRIFFID system, a comprehensive technical framework that integrates unmanned ground and aerial vehicles with advanced artificial intelligence functionalities to enhance disaster response capabilities across wildfires, urban floods, and post-earthquake search and rescue missions. By leveraging state-of-the-art autonomous navigation, semantic perception, and human-robot interaction technologies, TRIFFID provides a sophisticated system composed of the following key components: hybrid robotic platform, centralized ground station, custom communication infrastructure, and smartphone application. The defined research and development activities demonstrate how deep neural networks, knowledge graphs, and multimodal information fusion can enable robots to autonomously navigate and analyze disaster environments, reducing personnel risks and accelerating response times. The proposed system enhances emergency response teams by providing advanced mission planning, safety monitoring, and adaptive task execution capabilities. Moreover, it ensures real-time situational awareness and operational support in complex and risky situations, facilitating rapid and precise information collection and coordinated actions.
2502.09387
Truth Knows No Language: Evaluating Truthfulness Beyond English
cs.CL cs.AI cs.CY
We introduce a professionally translated extension of the TruthfulQA benchmark designed to evaluate truthfulness in Basque, Catalan, Galician, and Spanish. Truthfulness evaluations of large language models (LLMs) have primarily been conducted in English. However, the ability of LLMs to maintain truthfulness across languages remains under-explored. Our study evaluates 12 state-of-the-art open LLMs, comparing base and instruction-tuned models using human evaluation, multiple-choice metrics, and LLM-as-a-Judge scoring. Our findings reveal that, while LLMs perform best in English and worst in Basque (the lowest-resourced language), overall truthfulness discrepancies across languages are smaller than anticipated. Furthermore, we show that LLM-as-a-Judge correlates more closely with human judgments than multiple-choice metrics, and that informativeness plays a critical role in truthfulness assessment. Our results also indicate that machine translation provides a viable approach for extending truthfulness benchmarks to additional languages, offering a scalable alternative to professional translation. Finally, we observe that universal knowledge questions are better handled across languages than context- and time-dependent ones, highlighting the need for truthfulness evaluations that account for cultural and temporal variability. Dataset and code are publicly available under open licenses.
2502.09389
S$^2$-Diffusion: Generalizing from Instance-level to Category-level Skills in Robot Manipulation
cs.RO cs.AI
Recent advances in skill learning has propelled robot manipulation to new heights by enabling it to learn complex manipulation tasks from a practical number of demonstrations. However, these skills are often limited to the particular action, object, and environment \textit{instances} that are shown in the training data, and have trouble transferring to other instances of the same category. In this work we present an open-vocabulary Spatial-Semantic Diffusion policy (S$^2$-Diffusion) which enables generalization from instance-level training data to category-level, enabling skills to be transferable between instances of the same category. We show that functional aspects of skills can be captured via a promptable semantic module combined with a spatial representation. We further propose leveraging depth estimation networks to allow the use of only a single RGB camera. Our approach is evaluated and compared on a diverse number of robot manipulation tasks, both in simulation and in the real world. Our results show that S$^2$-Diffusion is invariant to changes in category-irrelevant factors as well as enables satisfying performance on other instances within the same category, even if it was not trained on that specific instance. Full videos of all real-world experiments are available in the supplementary material.
2502.09390
SQuARE: Sequential Question Answering Reasoning Engine for Enhanced Chain-of-Thought in Large Language Models
cs.CL cs.AI cs.LG
In the rapidly evolving field of Natural Language Processing, Large Language Models (LLMs) are tasked with increasingly complex reasoning challenges. Traditional methods like chain-of-thought prompting have shown promise but often fall short in fully leveraging a model's reasoning capabilities. This paper introduces SQuARE (Sequential Question Answering Reasoning Engine), a novel prompting technique designed to improve reasoning through a self-interrogation paradigm. Building upon CoT frameworks, SQuARE prompts models to generate and resolve multiple auxiliary questions before tackling the main query, promoting a more thorough exploration of various aspects of a topic. Our expansive evaluations, conducted with Llama 3 and GPT-4o models across multiple question-answering datasets, demonstrate that SQuARE significantly surpasses traditional CoT prompts and existing rephrase-and-respond methods. By systematically decomposing queries, SQuARE advances LLM capabilities in reasoning tasks. The code is publicly available at https://github.com/IntelLabs/RAG-FiT/tree/square.
2502.09393
Generalizable Reinforcement Learning with Biologically Inspired Hyperdimensional Occupancy Grid Maps for Exploration and Goal-Directed Path Planning
cs.RO cs.NE
Real-time autonomous systems utilize multi-layer computational frameworks to perform critical tasks such as perception, goal finding, and path planning. Traditional methods implement perception using occupancy grid mapping (OGM), segmenting the environment into discretized cells with probabilistic information. This classical approach is well-established and provides a structured input for downstream processes like goal finding and path planning algorithms. Recent approaches leverage a biologically inspired mathematical framework known as vector symbolic architectures (VSA), commonly known as hyperdimensional computing, to perform probabilistic OGM in hyperdimensional space. This approach, VSA-OGM, provides native compatibility with spiking neural networks, positioning VSA-OGM as a potential neuromorphic alternative to conventional OGM. However, for large-scale integration, it is essential to assess the performance implications of VSA-OGM on downstream tasks compared to established OGM methods. This study examines the efficacy of VSA-OGM against a traditional OGM approach, Bayesian Hilbert Maps (BHM), within reinforcement learning based goal finding and path planning frameworks, across a controlled exploration environment and an autonomous driving scenario inspired by the F1-Tenth challenge. Our results demonstrate that VSA-OGM maintains comparable learning performance across single and multi-scenario training configurations while improving performance on unseen environments by approximately 47%. These findings highlight the increased generalizability of policy networks trained with VSA-OGM over BHM, reinforcing its potential for real-world deployment in diverse environments.
2502.09395
Robot Pouring: Identifying Causes of Spillage and Selecting Alternative Action Parameters Using Probabilistic Actual Causation
cs.RO cs.LG
In everyday life, we perform tasks (e.g., cooking or cleaning) that involve a large variety of objects and goals. When confronted with an unexpected or unwanted outcome, we take corrective actions and try again until achieving the desired result. The reasoning performed to identify a cause of the observed outcome and to select an appropriate corrective action is a crucial aspect of human reasoning for successful task execution. Central to this reasoning is the assumption that a factor is responsible for producing the observed outcome. In this paper, we investigate the use of probabilistic actual causation to determine whether a factor is the cause of an observed undesired outcome. Furthermore, we show how the actual causation probabilities can be used to find alternative actions to change the outcome. We apply the probabilistic actual causation analysis to a robot pouring task. When spillage occurs, the analysis indicates whether a task parameter is the cause and how it should be changed to avoid spillage. The analysis requires a causal graph of the task and the corresponding conditional probability distributions. To fulfill these requirements, we perform a complete causal modeling procedure (i.e., task analysis, definition of variables, determination of the causal graph structure, and estimation of conditional probability distributions) using data from a realistic simulation of the robot pouring task, covering a large combinatorial space of task parameters. Based on the results, we discuss the implications of the variables' representation and how the alternative actions suggested by the actual causation analysis would compare to the alternative solutions proposed by a human observer. The practical use of the analysis of probabilistic actual causation to select alternative action parameters is demonstrated.
2502.09396
A hierarchical approach for assessing the vulnerability of tree-based classification models to membership inference attack
cs.LG cs.CR
Machine learning models can inadvertently expose confidential properties of their training data, making them vulnerable to membership inference attacks (MIA). While numerous evaluation methods exist, many require computationally expensive processes, such as training multiple shadow models. This article presents two new complementary approaches for efficiently identifying vulnerable tree-based models: an ante-hoc analysis of hyperparameter choices and a post-hoc examination of trained model structure. While these new methods cannot certify whether a model is safe from MIA, they provide practitioners with a means to significantly reduce the number of models that need to undergo expensive MIA assessment through a hierarchical filtering approach. More specifically, it is shown that the rank order of disclosure risk for different hyperparameter combinations remains consistent across datasets, enabling the development of simple, human-interpretable rules for identifying relatively high-risk models before training. While this ante-hoc analysis cannot determine absolute safety since this also depends on the specific dataset, it allows the elimination of unnecessarily risky configurations during hyperparameter tuning. Additionally, computationally inexpensive structural metrics serve as indicators of MIA vulnerability, providing a second filtering stage to identify risky models after training but before conducting expensive attacks. Empirical results show that hyperparameter-based risk prediction rules can achieve high accuracy in predicting the most at risk combinations of hyperparameters across different tree-based model types, while requiring no model training. Moreover, target model accuracy is not seen to correlate with privacy risk, suggesting opportunities to optimise model configurations for both performance and privacy.
2502.09411
ImageRAG: Dynamic Image Retrieval for Reference-Guided Image Generation
cs.CV cs.GR
Diffusion models enable high-quality and diverse visual content synthesis. However, they struggle to generate rare or unseen concepts. To address this challenge, we explore the usage of Retrieval-Augmented Generation (RAG) with image generation models. We propose ImageRAG, a method that dynamically retrieves relevant images based on a given text prompt, and uses them as context to guide the generation process. Prior approaches that used retrieved images to improve generation, trained models specifically for retrieval-based generation. In contrast, ImageRAG leverages the capabilities of existing image conditioning models, and does not require RAG-specific training. Our approach is highly adaptable and can be applied across different model types, showing significant improvement in generating rare and fine-grained concepts using different base models. Our project page is available at: https://rotem-shalev.github.io/ImageRAG
2502.09416
Rethinking Evaluation Metrics for Grammatical Error Correction: Why Use a Different Evaluation Process than Human?
cs.CL
One of the goals of automatic evaluation metrics in grammatical error correction (GEC) is to rank GEC systems such that it matches human preferences. However, current automatic evaluations are based on procedures that diverge from human evaluation. Specifically, human evaluation derives rankings by aggregating sentence-level relative evaluation results, e.g., pairwise comparisons, using a rating algorithm, whereas automatic evaluation averages sentence-level absolute scores to obtain corpus-level scores, which are then sorted to determine rankings. In this study, we propose an aggregation method for existing automatic evaluation metrics which aligns with human evaluation methods to bridge this gap. We conducted experiments using various metrics, including edit-based metrics, $n$-gram based metrics, and sentence-level metrics, and show that resolving the gap improves results for the most of metrics on the SEEDA benchmark. We also found that even BERT-based metrics sometimes outperform the metrics of GPT-4. We publish our unified implementation of the metrics and meta-evaluations.
2502.09417
A Survey of Reinforcement Learning for Optimization in Automation
cs.LG cs.AI cs.NE cs.RO cs.SY eess.SY
Reinforcement Learning (RL) has become a critical tool for optimization challenges within automation, leading to significant advancements in several areas. This review article examines the current landscape of RL within automation, with a particular focus on its roles in manufacturing, energy systems, and robotics. It discusses state-of-the-art methods, major challenges, and upcoming avenues of research within each sector, highlighting RL's capacity to solve intricate optimization challenges. The paper reviews the advantages and constraints of RL-driven optimization methods in automation. It points out prevalent challenges encountered in RL optimization, including issues related to sample efficiency and scalability; safety and robustness; interpretability and trustworthiness; transfer learning and meta-learning; and real-world deployment and integration. It further explores prospective strategies and future research pathways to navigate these challenges. Additionally, the survey includes a comprehensive list of relevant research papers, making it an indispensable guide for scholars and practitioners keen on exploring this domain.
2502.09419
On multi-token prediction for efficient LLM inference
cs.CL cs.LG
We systematically investigate multi-token prediction (MTP) capabilities within LLMs pre-trained for next-token prediction (NTP). We first show that such models inherently possess MTP capabilities via numerical marginalization over intermediate token probabilities, though performance is data-dependent and improves with model scale. Furthermore, we explore the challenges of integrating MTP heads into frozen LLMs and find that their hidden layers are strongly specialized for NTP, making adaptation non-trivial. Finally, we show that while joint training of MTP heads with the backbone improves performance, it cannot fully overcome this barrier, prompting further research in this direction. Our findings provide a deeper understanding of MTP applied to pretrained LLMs, informing strategies for accelerating inference through parallel token prediction.
2502.09423
Transformer-Enhanced Variational Autoencoder for Crystal Structure Prediction
cond-mat.mtrl-sci cs.AI
Crystal structure forms the foundation for understanding the physical and chemical properties of materials. Generative models have emerged as a new paradigm in crystal structure prediction(CSP), however, accurately capturing key characteristics of crystal structures, such as periodicity and symmetry, remains a significant challenge. In this paper, we propose a Transformer-Enhanced Variational Autoencoder for Crystal Structure Prediction (TransVAE-CSP), who learns the characteristic distribution space of stable materials, enabling both the reconstruction and generation of crystal structures. TransVAE-CSP integrates adaptive distance expansion with irreducible representation to effectively capture the periodicity and symmetry of crystal structures, and the encoder is a transformer network based on an equivariant dot product attention mechanism. Experimental results on the carbon_24, perov_5, and mp_20 datasets demonstrate that TransVAE-CSP outperforms existing methods in structure reconstruction and generation tasks under various modeling metrics, offering a powerful tool for crystal structure design and optimization.
2502.09425
A 3D Facial Reconstruction Evaluation Methodology: Comparing Smartphone Scans with Deep Learning Based Methods Using Geometry and Morphometry Criteria
cs.CV
Three-dimensional (3D) facial shape analysis has gained interest due to its potential clinical applications. However, the high cost of advanced 3D facial acquisition systems limits their widespread use, driving the development of low-cost acquisition and reconstruction methods. This study introduces a novel evaluation methodology that goes beyond traditional geometry-based benchmarks by integrating morphometric shape analysis techniques, providing a statistical framework for assessing facial morphology preservation. As a case study, we compare smartphone-based 3D scans with state-of-the-art deep learning reconstruction methods from 2D images, using high-end stereophotogrammetry models as ground truth. This methodology enables a quantitative assessment of global and local shape differences, offering a biologically meaningful validation approach for low-cost 3D facial acquisition and reconstruction techniques.
2502.09431
On Usage of Non-Volatile Memory as Primary Storage for Database Management Systems
cs.DB
This paper explores the implications of employing non-volatile memory (NVM) as primary storage for a data base management system (DBMS). We investigate the modifications necessary to be applied on top of a traditional relational DBMS to take advantage of NVM features. As a case study, we modify the storage engine (SE) of PostgreSQL enabling efficient use of NVM hardware. We detail the necessary changes and challenges such modifications entail and evaluate them using a comprehensive emulation platform. Results indicate that our modified SE reduces query execution time by up to 45% and 13% when compared to disk and NVM storage, with average reductions of 19% and 4%, respectively. Detailed analysis of these results shows that while our modified SE is able to access data more efficiently, data is not close to the processing units when needed for processing, incurring long latency misses that hinder the performance. To solve this, we develop a general purpose library that employs helper threads to prefetch data from NVM hardware via a simple API. Our library further improves query execution time for our modified SE when compared to disk and NVM storage by up to 54% and 17%, with average reductions of 23% and 8%, respectively.
2502.09432
Dual Formulation for Non-Rectangular Lp Robust Markov Decision Processes
cs.AI cs.LG
We study robust Markov decision processes (RMDPs) with non-rectangular uncertainty sets, which capture interdependencies across states unlike traditional rectangular models. While non-rectangular robust policy evaluation is generally NP-hard, even in approximation, we identify a powerful class of $L_p$-bounded uncertainty sets that avoid these complexity barriers due to their structural simplicity. We further show that this class can be decomposed into infinitely many \texttt{sa}-rectangular $L_p$-bounded sets and leverage its structural properties to derive a novel dual formulation for $L_p$ RMDPs. This formulation provides key insights into the adversary's strategy and enables the development of the first robust policy evaluation algorithms for non-rectangular RMDPs. Empirical results demonstrate that our approach significantly outperforms brute-force methods, establishing a promising foundation for future investigation into non-rectangular robust MDPs.
2502.09434
Redistribute Ensemble Training for Mitigating Memorization in Diffusion Models
cs.CV
Diffusion models, known for their tremendous ability to generate high-quality samples, have recently raised concerns due to their data memorization behavior, which poses privacy risks. Recent methods for memory mitigation have primarily addressed the issue within the context of the text modality in cross-modal generation tasks, restricting their applicability to specific conditions. In this paper, we propose a novel method for diffusion models from the perspective of visual modality, which is more generic and fundamental for mitigating memorization. Directly exposing visual data to the model increases memorization risk, so we design a framework where models learn through proxy model parameters instead. Specially, the training dataset is divided into multiple shards, with each shard training a proxy model, then aggregated to form the final model. Additionally, practical analysis of training losses illustrates that the losses for easily memorable images tend to be obviously lower. Thus, we skip the samples with abnormally low loss values from the current mini-batch to avoid memorizing. However, balancing the need to skip memorization-prone samples while maintaining sufficient training data for high-quality image generation presents a key challenge. Thus, we propose IET-AGC+, which redistributes highly memorizable samples between shards, to mitigate these samples from over-skipping. Furthermore, we dynamically augment samples based on their loss values to further reduce memorization. Extensive experiments and analysis on four datasets show that our method successfully reduces memory capacity while maintaining performance. Moreover, we fine-tune the pre-trained diffusion models, e.g., Stable Diffusion, and decrease the memorization score by 46.7\%, demonstrating the effectiveness of our method. Code is available in: https://github.com/liuxiao-guan/IET_AGC.
2502.09436
Variable Stiffness for Robust Locomotion through Reinforcement Learning
cs.RO cs.AI
Reinforcement-learned locomotion enables legged robots to perform highly dynamic motions but often accompanies time-consuming manual tuning of joint stiffness. This paper introduces a novel control paradigm that integrates variable stiffness into the action space alongside joint positions, enabling grouped stiffness control such as per-joint stiffness (PJS), per-leg stiffness (PLS) and hybrid joint-leg stiffness (HJLS). We show that variable stiffness policies, with grouping in per-leg stiffness (PLS), outperform position-based control in velocity tracking and push recovery. In contrast, HJLS excels in energy efficiency. Furthermore, our method showcases robust walking behaviour on diverse outdoor terrains by sim-to-real transfer, although the policy is sorely trained on a flat floor. Our approach simplifies design by eliminating per-joint stiffness tuning while keeping competitive results with various metrics.
2502.09443
Relational Conformal Prediction for Correlated Time Series
cs.LG cs.AI
We address the problem of uncertainty quantification in time series forecasting by exploiting observations at correlated sequences. Relational deep learning methods leveraging graph representations are among the most effective tools for obtaining point estimates from spatiotemporal data and correlated time series. However, the problem of exploiting relational structures to estimate the uncertainty of such predictions has been largely overlooked in the same context. To this end, we propose a novel distribution-free approach based on the conformal prediction framework and quantile regression. Despite the recent applications of conformal prediction to sequential data, existing methods operate independently on each target time series and do not account for relationships among them when constructing the prediction interval. We fill this void by introducing a novel conformal prediction method based on graph deep learning operators. Our method, named Conformal Relational Prediction (CoRel), does not require the relational structure (graph) to be known as a prior and can be applied on top of any pre-trained time series predictor. Additionally, CoRel includes an adaptive component to handle non-exchangeable data and changes in the input time series. Our approach provides accurate coverage and archives state-of-the-art uncertainty quantification in relevant benchmarks.
2502.09445
A Differentiable Rank-Based Objective For Better Feature Learning
stat.ML cs.LG
In this paper, we leverage existing statistical methods to better understand feature learning from data. We tackle this by modifying the model-free variable selection method, Feature Ordering by Conditional Independence (FOCI), which is introduced in \cite{azadkia2021simple}. While FOCI is based on a non-parametric coefficient of conditional dependence, we introduce its parametric, differentiable approximation. With this approximate coefficient of correlation, we present a new algorithm called difFOCI, which is applicable to a wider range of machine learning problems thanks to its differentiable nature and learnable parameters. We present difFOCI in three contexts: (1) as a variable selection method with baseline comparisons to FOCI, (2) as a trainable model parametrized with a neural network, and (3) as a generic, widely applicable neural network regularizer, one that improves feature learning with better management of spurious correlations. We evaluate difFOCI on increasingly complex problems ranging from basic variable selection in toy examples to saliency map comparisons in convolutional networks. We then show how difFOCI can be incorporated in the context of fairness to facilitate classifications without relying on sensitive data.
2502.09446
Drivers of cooperation in social dilemmas on higher-order networks
physics.soc-ph cs.GT cs.SI q-bio.PE
Understanding cooperation in social dilemmas requires models that capture the complexity of real-world interactions. While network frameworks have provided valuable insights to model the evolution of cooperation, they are unable to encode group interactions properly. Here, we introduce a general higher-order network framework for multi-player games on structured populations. Our model considers multi-dimensional strategies, based on the observation that social behaviours are affected by the size of the group interaction. We investigate dynamical and structural coupling between different orders of interactions, revealing the crucial role of nested multilevel interactions, and showing how such features can enhance cooperation beyond the limit of traditional models with uni-dimensional strategies. Our work identifies the key drivers promoting cooperative behaviour commonly observed in real-world group social dilemmas.
2502.09447
Pixel-Level Reasoning Segmentation via Multi-turn Conversations
cs.CV cs.CL
Existing visual perception systems focus on region-level segmentation in single-turn dialogues, relying on complex and explicit query instructions. Such systems cannot reason at the pixel level and comprehend dynamic user intent that changes over interaction. Our work tackles this issue by introducing a novel task, Pixel-level Reasoning Segmentation (Pixel-level RS) based on multi-turn conversations, tracking evolving user intent via multi-turn interactions for fine-grained segmentation. To establish a benchmark for this novel task, we build a Pixel-level ReasonIng Segmentation Dataset Based on Multi-Turn Conversations (PRIST), comprising 24k utterances from 8.3k multi-turn conversational scenarios with segmentation targets. Building on PRIST, we further propose MIRAS, a Multi-turn Interactive ReAsoning Segmentation framework, integrates pixel-level segmentation with robust multi-turn conversation understanding, generating pixel-grounded explanations aligned with user intent. The PRIST dataset and MIRSA framework fill the gap in pixel-level reasoning segmentation. Experimental results on the PRIST dataset demonstrate that our method outperforms current segmentation-specific baselines in terms of segmentation and LLM-based reasoning metrics. The code and data are available at: https://github.com/ccccai239/PixelRIST.
2502.09449
Spiking Neural Networks for Temporal Processing: Status Quo and Future Prospects
cs.NE
Temporal processing is fundamental for both biological and artificial intelligence systems, as it enables the comprehension of dynamic environments and facilitates timely responses. Spiking Neural Networks (SNNs) excel in handling such data with high efficiency, owing to their rich neuronal dynamics and sparse activity patterns. Given the recent surge in the development of SNNs, there is an urgent need for a comprehensive evaluation of their temporal processing capabilities. In this paper, we first conduct an in-depth assessment of commonly used neuromorphic benchmarks, revealing critical limitations in their ability to evaluate the temporal processing capabilities of SNNs. To bridge this gap, we further introduce a benchmark suite consisting of three temporal processing tasks characterized by rich temporal dynamics across multiple timescales. Utilizing this benchmark suite, we perform a thorough evaluation of recently introduced SNN approaches to elucidate the current status of SNNs in temporal processing. Our findings indicate significant advancements in recently developed spiking neuron models and neural architectures regarding their temporal processing capabilities, while also highlighting a performance gap in handling long-range dependencies when compared to state-of-the-art non-spiking models. Finally, we discuss the key challenges and outline potential avenues for future research.
2502.09457
The Multilingual Mind : A Survey of Multilingual Reasoning in Language Models
cs.CL
While reasoning and multilingual capabilities in Language Models (LMs) have achieved remarkable progress in recent years, their integration into a unified paradigm, multilingual reasoning, is at a nascent stage. Multilingual reasoning requires language models to handle logical reasoning across languages while addressing misalignment, biases, and challenges in low-resource settings. This survey provides the first in-depth review of multilingual reasoning in LMs. In this survey, we provide a systematic overview of existing methods that leverage LMs for multilingual reasoning, specifically outlining the challenges, motivations, and foundational aspects of applying language models to reason across diverse languages. We provide an overview of the standard data resources used for training multilingual reasoning in LMs and the evaluation benchmarks employed to assess their multilingual capabilities. Next, we analyze various state-of-the-art methods and their performance on these benchmarks. Finally, we explore future research opportunities to improve multilingual reasoning in LMs, focusing on enhancing their ability to handle diverse languages and complex reasoning tasks.
2502.09460
Metamorphic Testing for Pose Estimation Systems
cs.SE cs.AI cs.CV
Pose estimation systems are used in a variety of fields, from sports analytics to livestock care. Given their potential impact, it is paramount to systematically test their behaviour and potential for failure. This is a complex task due to the oracle problem and the high cost of manual labelling necessary to build ground truth keypoints. This problem is exacerbated by the fact that different applications require systems to focus on different subjects (e.g., human versus animal) or landmarks (e.g., only extremities versus whole body and face), which makes labelled test data rarely reusable. To combat these problems we propose MET-POSE, a metamorphic testing framework for pose estimation systems that bypasses the need for manual annotation while assessing the performance of these systems under different circumstances. MET-POSE thus allows users of pose estimation systems to assess the systems in conditions that more closely relate to their application without having to label an ad-hoc test dataset or rely only on available datasets, which may not be adapted to their application domain. While we define MET-POSE in general terms, we also present a non-exhaustive list of metamorphic rules that represent common challenges in computer vision applications, as well as a specific way to evaluate these rules. We then experimentally show the effectiveness of MET-POSE by applying it to Mediapipe Holistic, a state of the art human pose estimation system, with the FLIC and PHOENIX datasets. With these experiments, we outline numerous ways in which the outputs of MET-POSE can uncover faults in pose estimation systems at a similar or higher rate than classic testing using hand labelled data, and show that users can tailor the rule set they use to the faults and level of accuracy relevant to their application.
2502.09471
Wholly-WOOD: Wholly Leveraging Diversified-quality Labels for Weakly-supervised Oriented Object Detection
cs.CV cs.AI
Accurately estimating the orientation of visual objects with compact rotated bounding boxes (RBoxes) has become a prominent demand, which challenges existing object detection paradigms that only use horizontal bounding boxes (HBoxes). To equip the detectors with orientation awareness, supervised regression/classification modules have been introduced at the high cost of rotation annotation. Meanwhile, some existing datasets with oriented objects are already annotated with horizontal boxes or even single points. It becomes attractive yet remains open for effectively utilizing weaker single point and horizontal annotations to train an oriented object detector (OOD). We develop Wholly-WOOD, a weakly-supervised OOD framework, capable of wholly leveraging various labeling forms (Points, HBoxes, RBoxes, and their combination) in a unified fashion. By only using HBox for training, our Wholly-WOOD achieves performance very close to that of the RBox-trained counterpart on remote sensing and other areas, significantly reducing the tedious efforts on labor-intensive annotation for oriented objects. The source codes are available at https://github.com/VisionXLab/whollywood (PyTorch-based) and https://github.com/VisionXLab/whollywood-jittor (Jittor-based).
2502.09473
Learning to Predict Global Atrial Fibrillation Dynamics from Sparse Measurements
cs.LG eess.SP
Catheter ablation of Atrial Fibrillation (AF) consists of a one-size-fits-all treatment with limited success in persistent AF. This may be due to our inability to map the dynamics of AF with the limited resolution and coverage provided by sequential contact mapping catheters, preventing effective patient phenotyping for personalised, targeted ablation. Here we introduce FibMap, a graph recurrent neural network model that reconstructs global AF dynamics from sparse measurements. Trained and validated on 51 non-contact whole atria recordings, FibMap reconstructs whole atria dynamics from 10% surface coverage, achieving a 210% lower mean absolute error and an order of magnitude higher performance in tracking phase singularities compared to baseline methods. Clinical utility of FibMap is demonstrated on real-world contact mapping recordings, achieving reconstruction fidelity comparable to non-contact mapping. FibMap's state-spaces and patient-specific parameters offer insights for electrophenotyping AF. Integrating FibMap into clinical practice could enable personalised AF care and improve outcomes.
2502.09477
DiffRenderGAN: Addressing Training Data Scarcity in Deep Segmentation Networks for Quantitative Nanomaterial Analysis through Differentiable Rendering and Generative Modelling
cond-mat.mtrl-sci cs.CV cs.LG
Nanomaterials exhibit distinctive properties governed by parameters such as size, shape, and surface characteristics, which critically influence their applications and interactions across technological, biological, and environmental contexts. Accurate quantification and understanding of these materials are essential for advancing research and innovation. In this regard, deep learning segmentation networks have emerged as powerful tools that enable automated insights and replace subjective methods with precise quantitative analysis. However, their efficacy depends on representative annotated datasets, which are challenging to obtain due to the costly imaging of nanoparticles and the labor-intensive nature of manual annotations. To overcome these limitations, we introduce DiffRenderGAN, a novel generative model designed to produce annotated synthetic data. By integrating a differentiable renderer into a Generative Adversarial Network (GAN) framework, DiffRenderGAN optimizes textural rendering parameters to generate realistic, annotated nanoparticle images from non-annotated real microscopy images. This approach reduces the need for manual intervention and enhances segmentation performance compared to existing synthetic data methods by generating diverse and realistic data. Tested on multiple ion and electron microscopy cases, including titanium dioxide (TiO$_2$), silicon dioxide (SiO$_2$)), and silver nanowires (AgNW), DiffRenderGAN bridges the gap between synthetic and real data, advancing the quantification and understanding of complex nanomaterial systems.
2502.09479
Assessing Generative AI value in a public sector context: evidence from a field experiment
q-fin.GN cs.LG econ.GN q-fin.EC
The emergence of Generative AI (Gen AI) has motivated an interest in understanding how it could be used to enhance productivity across various tasks. We add to research results for the performance impact of Gen AI on complex knowledge-based tasks in a public sector setting. In a pre-registered experiment, after establishing a baseline level of performance, we find mixed evidence for two types of composite tasks related to document understanding and data analysis. For the Documents task, the treatment group using Gen AI had a 17% improvement in answer quality scores (as judged by human evaluators) and a 34% improvement in task completion time compared to a control group. For the Data task, we find the Gen AI treatment group experienced a 12% reduction in quality scores and no significant difference in mean completion time compared to the control group. These results suggest that the benefits of Gen AI may be task and potentially respondent dependent. We also discuss field notes and lessons learned, as well as supplementary insights from a post-trial survey and feedback workshop with participants.
2502.09482
Standardisation of Convex Ultrasound Data Through Geometric Analysis and Augmentation
cs.CV
The application of ultrasound in healthcare has seen increased diversity and importance. Unlike other medical imaging modalities, ultrasound research and development has historically lagged, particularly in the case of applications with data-driven algorithms. A significant issue with ultrasound is the extreme variability of the images, due to the number of different machines available and the possible combination of parameter settings. One outcome of this is the lack of standardised and benchmarking ultrasound datasets. The method proposed in this article is an approach to alleviating this issue of disorganisation. For this purpose, the issue of ultrasound data sparsity is examined and a novel perspective, approach, and solution is proposed; involving the extraction of the underlying ultrasound plane within the image and representing it using annulus sector geometry. An application of this methodology is proposed, which is the extraction of scan lines and the linearisation of convex planes. Validation of the robustness of the proposed method is performed on both private and public data. The impact of deformation and the invertibility of augmentation using the estimated annulus sector parameters is also studied. Keywords: Ultrasound, Annulus Sector, Augmentation, Linearisation.
2502.09484
PenTest++: Elevating Ethical Hacking with AI and Automation
cs.CR cs.AI
Traditional ethical hacking relies on skilled professionals and time-intensive command management, which limits its scalability and efficiency. To address these challenges, we introduce PenTest++, an AI-augmented system that integrates automation with generative AI (GenAI) to optimise ethical hacking workflows. Developed in a controlled virtual environment, PenTest++ streamlines critical penetration testing tasks, including reconnaissance, scanning, enumeration, exploitation, and documentation, while maintaining a modular and adaptable design. The system balances automation with human oversight, ensuring informed decision-making at key stages, and offers significant benefits such as enhanced efficiency, scalability, and adaptability. However, it also raises ethical considerations, including privacy concerns and the risks of AI-generated inaccuracies (hallucinations). This research underscores the potential of AI-driven systems like PenTest++ to complement human expertise in cybersecurity by automating routine tasks, enabling professionals to focus on strategic decision-making. By incorporating robust ethical safeguards and promoting ongoing refinement, PenTest++ demonstrates how AI can be responsibly harnessed to address operational and ethical challenges in the evolving cybersecurity landscape.
2502.09487
Objective quantification of mood states using large language models
cs.CL cs.AI cs.LG
Emotional states influence human behaviour and cognition, leading to diverse thought trajectories. Similarly, Large Language Models (LLMs) showcase an excellent level of response consistency across wide-ranging contexts (prompts). We leverage these parallels to establish a framework for quantifying mental states. Our approach utilises self-report questionnaires that reliably assess these states due to their inherent sensitivity to patterns of co-occurring responses. Specifically, we recruited a large sample of participants (N=422) to investigate how well an LLM (Mistral-7B-OpenOrca) quantifies a heterogenous set of depressive mood states measured with participants' open-ended responses to a depression questionnaire. We show LLM responses to held-out multiple-choice questions, given participants' open-ended answers, correlate strongly (r: 0.52-0.84) with true questionnaire scores, demonstrating LLM's generalisation from mood representations. We explore a link between these representations and factor analysis. Using ridge regression, we find depression-related subspaces within LLM hidden states. We show these subspaces to be predictive of participants' "Depression" and "Somatic & Emotional Distress" factor scores, as well as suicidality severity. Overall, LLMs can provide quantitative measures of mental states. The reliability of these hinges upon how informative the questions we ask participants are. Used correctly, this approach could supplement mental state assessment in a variety of settings.
2502.09490
Inverse Design with Dynamic Mode Decomposition
cs.LG cs.SY eess.SY math.DS math.OC physics.flu-dyn
We introduce a computationally efficient method for the automation of inverse design in science and engineering. Based on simple least-square regression, the underlying dynamic mode decomposition algorithm can be used to construct a low-rank subspace spanning multiple experiments in parameter space. The proposed inverse design dynamic mode composition (ID-DMD) algorithm leverages the computed low-dimensional subspace to enable fast digital design and optimization on laptop-level computing, including the potential to prescribe the dynamics themselves. Moreover, the method is robust to noise, physically interpretable, and can provide uncertainty quantification metrics. The architecture can also efficiently scale to large-scale design problems using randomized algorithms in the ID-DMD. The simplicity of the method and its implementation are highly attractive in practice, and the ID-DMD has been demonstrated to be an order of magnitude more accurate than competing methods while simultaneously being 3-5 orders faster on challenging engineering design problems ranging from structural vibrations to fluid dynamics. Due to its speed, robustness, interpretability, and ease-of-use, ID-DMD in comparison with other leading machine learning methods represents a significant advancement in data-driven methods for inverse design and optimization, promising a paradigm shift in how to approach inverse design in practice.
2502.09494
Communicating Likelihoods with Normalising Flows
hep-ph cs.LG hep-ex physics.data-an
We present a machine-learning-based workflow to model an unbinned likelihood from its samples. A key advancement over existing approaches is the validation of the learned likelihood using rigorous statistical tests of the joint distribution, such as the Kolmogorov-Smirnov test of the joint distribution. Our method enables the reliable communication of experimental and phenomenological likelihoods for subsequent analyses. We demonstrate its effectiveness through three case studies in high-energy physics. To support broader adoption, we provide an open-source reference implementation, nabu.
2502.09495
Cracking the Code: Enhancing Development finance understanding with artificial intelligence
econ.GN cs.AI cs.LG q-fin.EC
Analyzing development projects is crucial for understanding donors aid strategies, recipients priorities, and to assess development finance capacity to adress development issues by on-the-ground actions. In this area, the Organisation for Economic Co-operation and Developments (OECD) Creditor Reporting System (CRS) dataset is a reference data source. This dataset provides a vast collection of project narratives from various sectors (approximately 5 million projects). While the OECD CRS provides a rich source of information on development strategies, it falls short in informing project purposes due to its reporting process based on donors self-declared main objectives and pre-defined industrial sectors. This research employs a novel approach that combines Machine Learning (ML) techniques, specifically Natural Language Processing (NLP), an innovative Python topic modeling technique called BERTopic, to categorise (cluster) and label development projects based on their narrative descriptions. By revealing existing yet hidden topics of development finance, this application of artificial intelligence enables a better understanding of donor priorities and overall development funding and provides methods to analyse public and private projects narratives.
2502.09496
On Agnostic PAC Learning in the Small Error Regime
cs.LG stat.ML
Binary classification in the classic PAC model exhibits a curious phenomenon: Empirical Risk Minimization (ERM) learners are suboptimal in the realizable case yet optimal in the agnostic case. Roughly speaking, this owes itself to the fact that non-realizable distributions $\mathcal{D}$ are simply more difficult to learn than realizable distributions -- even when one discounts a learner's error by $\mathrm{err}(h^*_{\mathcal{D}})$, the error of the best hypothesis in $\mathcal{H}$ for $\mathcal{D}$. Thus, optimal agnostic learners are permitted to incur excess error on (easier-to-learn) distributions $\mathcal{D}$ for which $\tau = \mathrm{err}(h^*_{\mathcal{D}})$ is small. Recent work of Hanneke, Larsen, and Zhivotovskiy (FOCS `24) addresses this shortcoming by including $\tau$ itself as a parameter in the agnostic error term. In this more fine-grained model, they demonstrate tightness of the error lower bound $\tau + \Omega \left(\sqrt{\frac{\tau (d + \log(1 / \delta))}{m}} + \frac{d + \log(1 / \delta)}{m} \right)$ in a regime where $\tau > d/m$, and leave open the question of whether there may be a higher lower bound when $\tau \approx d/m$, with $d$ denoting $\mathrm{VC}(\mathcal{H})$. In this work, we resolve this question by exhibiting a learner which achieves error $c \cdot \tau + O \left(\sqrt{\frac{\tau (d + \log(1 / \delta))}{m}} + \frac{d + \log(1 / \delta)}{m} \right)$ for a constant $c \leq 2.1$, thus matching the lower bound when $\tau \approx d/m$. Further, our learner is computationally efficient and is based upon careful aggregations of ERM classifiers, making progress on two other questions of Hanneke, Larsen, and Zhivotovskiy (FOCS `24). We leave open the interesting question of whether our approach can be refined to lower the constant from 2.1 to 1, which would completely settle the complexity of agnostic learning.
2502.09497
Improve LLM-based Automatic Essay Scoring with Linguistic Features
cs.CL cs.AI
Automatic Essay Scoring (AES) assigns scores to student essays, reducing the grading workload for instructors. Developing a scoring system capable of handling essays across diverse prompts is challenging due to the flexibility and diverse nature of the writing task. Existing methods typically fall into two categories: supervised feature-based approaches and large language model (LLM)-based methods. Supervised feature-based approaches often achieve higher performance but require resource-intensive training. In contrast, LLM-based methods are computationally efficient during inference but tend to suffer from lower performance. This paper combines these approaches by incorporating linguistic features into LLM-based scoring. Experimental results show that this hybrid method outperforms baseline models for both in-domain and out-of-domain writing prompts.
2502.09500
Eidetic Learning: an Efficient and Provable Solution to Catastrophic Forgetting
cs.LG
Catastrophic forgetting -- the phenomenon of a neural network learning a task t1 and losing the ability to perform it after being trained on some other task t2 -- is a long-standing problem for neural networks [McCloskey and Cohen, 1989]. We present a method, Eidetic Learning, that provably solves catastrophic forgetting. A network trained with Eidetic Learning -- here, an EideticNet -- requires no rehearsal or replay. We consider successive discrete tasks and show how at inference time an EideticNet automatically routes new instances without auxiliary task information. An EideticNet bears a family resemblance to the sparsely-gated Mixture-of-Experts layer Shazeer et al. [2016] in that network capacity is partitioned across tasks and the network itself performs data-conditional routing. An EideticNet is easy to implement and train, is efficient, and has time and space complexity linear in the number of parameters. The guarantee of our method holds for normalization layers of modern neural networks during both pre-training and fine-tuning. We show with a variety of network architectures and sets of tasks that EideticNets are immune to forgetting. While the practical benefits of EideticNets are substantial, we believe they can be benefit practitioners and theorists alike. The code for training EideticNets is available at https://github.com/amazon-science/eideticnet-training.
2502.09501
Prior-Constrained Association Learning for Fine-Grained Generalized Category Discovery
cs.CV
This paper addresses generalized category discovery (GCD), the task of clustering unlabeled data from potentially known or unknown categories with the help of labeled instances from each known category. Compared to traditional semi-supervised learning, GCD is more challenging because unlabeled data could be from novel categories not appearing in labeled data. Current state-of-the-art methods typically learn a parametric classifier assisted by self-distillation. While being effective, these methods do not make use of cross-instance similarity to discover class-specific semantics which are essential for representation learning and category discovery. In this paper, we revisit the association-based paradigm and propose a Prior-constrained Association Learning method to capture and learn the semantic relations within data. In particular, the labeled data from known categories provides a unique prior for the association of unlabeled data. Unlike previous methods that only adopts the prior as a pre or post-clustering refinement, we fully incorporate the prior into the association process, and let it constrain the association towards a reliable grouping outcome. The estimated semantic groups are utilized through non-parametric prototypical contrast to enhance the representation learning. A further combination of both parametric and non-parametric classification complements each other and leads to a model that outperforms existing methods by a significant margin. On multiple GCD benchmarks, we perform extensive experiments and validate the effectiveness of our proposed method.
2502.09502
Scalable First-order Method for Certifying Optimal k-Sparse GLMs
cs.LG math.OC
This paper investigates the problem of certifying optimality for sparse generalized linear models (GLMs), where sparsity is enforced through an $\ell_0$ cardinality constraint. While branch-and-bound (BnB) frameworks can certify optimality by pruning nodes using dual bounds, existing methods for computing these bounds are either computationally intensive or exhibit slow convergence, limiting their scalability to large-scale problems. To address this challenge, we propose a first-order proximal gradient algorithm designed to solve the perspective relaxation of the problem within a BnB framework. Specifically, we formulate the relaxed problem as a composite optimization problem and demonstrate that the proximal operator of the non-smooth component can be computed exactly in log-linear time complexity, eliminating the need to solve a computationally expensive second-order cone program. Furthermore, we introduce a simple restart strategy that enhances convergence speed while maintaining low per-iteration complexity. Extensive experiments on synthetic and real-world datasets show that our approach significantly accelerates dual bound computations and is highly effective in providing optimality certificates for large-scale problems.
2502.09503
AttentionSmithy: A Modular Framework for Rapid Transformer Development and Customization
cs.LG cs.AI
Transformer architectures have transformed AI applications but remain complex to customize for domain experts lacking low-level implementation expertise. We introduce AttentionSmithy, a modular software package that simplifies transformer innovation by breaking down key components into reusable building blocks: attention modules, feed-forward networks, normalization layers, and positional encodings. Users can rapidly prototype and evaluate transformer variants without extensive coding. Our framework supports four positional encoding strategies and integrates with neural architecture search for automated design. We validate AttentionSmithy by replicating the original transformer under resource constraints and optimizing translation performance by combining positional encodings. Additionally, we demonstrate its adaptability in gene-specific modeling, achieving over 95% accuracy in cell type classification. These case studies highlight AttentionSmithy's potential to accelerate research across diverse fields by removing framework implementation barriers.
2502.09507
When and How Does CLIP Enable Domain and Compositional Generalization?
cs.LG cs.CV
The remarkable generalization performance of contrastive vision-language models like CLIP is often attributed to the diversity of their training distributions. However, key questions remain unanswered: Can CLIP generalize to an entirely unseen domain when trained on a diverse mixture of domains (domain generalization)? Can it generalize to unseen classes within partially seen domains (compositional generalization)? What factors affect such generalization? To answer these questions, we trained CLIP models on systematically constructed training distributions with controlled domain diversity and object class exposure. Our experiments show that domain diversity is essential for both domain and compositional generalization, yet compositional generalization can be surprisingly weaker than domain generalization when the training distribution contains a suboptimal subset of the test domain. Through data-centric and mechanistic analyses, we find that successful generalization requires learning of shared representations already in intermediate layers and shared circuitry.
2502.09509
EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling
cs.LG
Latent generative models have emerged as a leading approach for high-quality image synthesis. These models rely on an autoencoder to compress images into a latent space, followed by a generative model to learn the latent distribution. We identify that existing autoencoders lack equivariance to semantic-preserving transformations like scaling and rotation, resulting in complex latent spaces that hinder generative performance. To address this, we propose EQ-VAE, a simple regularization approach that enforces equivariance in the latent space, reducing its complexity without degrading reconstruction quality. By finetuning pre-trained autoencoders with EQ-VAE, we enhance the performance of several state-of-the-art generative models, including DiT, SiT, REPA and MaskGIT, achieving a 7 speedup on DiT-XL/2 with only five epochs of SD-VAE fine-tuning. EQ-VAE is compatible with both continuous and discrete autoencoders, thus offering a versatile enhancement for a wide range of latent generative models. Project page and code: https://eq-vae.github.io/.
2502.09511
Diffusion Models for Molecules: A Survey of Methods and Tasks
cs.LG cs.AI cs.CE
Generative tasks about molecules, including but not limited to molecule generation, are crucial for drug discovery and material design, and have consistently attracted significant attention. In recent years, diffusion models have emerged as an impressive class of deep generative models, sparking extensive research and leading to numerous studies on their application to molecular generative tasks. Despite the proliferation of related work, there remains a notable lack of up-to-date and systematic surveys in this area. Particularly, due to the diversity of diffusion model formulations, molecular data modalities, and generative task types, the research landscape is challenging to navigate, hindering understanding and limiting the area's growth. To address this, this paper conducts a comprehensive survey of diffusion model-based molecular generative methods. We systematically review the research from the perspectives of methodological formulations, data modalities, and task types, offering a novel taxonomy. This survey aims to facilitate understanding and further flourishing development in this area. The relevant papers are summarized at: https://github.com/AzureLeon1/awesome-molecular-diffusion-models.
2502.09517
Coupled Rendezvous and Docking Maneuver control of satellite using Reinforcement learning-based Adaptive Fixed-Time Sliding Mode Controller
eess.SY cs.SY
Satellite dynamics in unknown environments are inherently uncertain due to factors such as varying gravitational fields, atmospheric drag, and unpredictable interactions with space debris or other celestial bodies. Traditional sliding mode controllers with fixed parameters often struggle to maintain optimal performance under these fluctuating conditions. Therefore, an adaptive controller is essential to address these challenges by continuously tuning its gains in real-time. In this paper, we have tuned the slopes of the Fixed-time Sliding surface adaptively using reinforcement learning for coupled rendezvous and docking maneuver of chaser satellite with the target satellite in an unknown space environment. The neural network model is used to determine the optimal gains of reaching law of the fixed-time sliding surface. We have assumed that we don't have an accurate model of the system so we have added noise in the tangent space instead of directly on the manifold to preserve the geometric structure of the system while ensuring mathematically consistent uncertainty propagation. The reinforcement learning is used as an approximator to represent the value function of the agent to estimate the dynamical model of the system using the Actor-Critic method. The proposed control algorithm integrates a neural network and a sliding mode controller in a cascade loop architecture, where the tracking error dynamically tunes the sliding surface gains. Global fixed-time stability of the closed-loop feedback system is proved within the Lyapunov framework. This comprehensive approach of fixed-time sliding mode controller using a Reinforcement Learning based ensures the completion of the mission efficiently while addressing the critical challenges posed by the uncertain environment. The simulation results presented support the claims made.
2502.09520
SQ-GAN: Semantic Image Communications Using Masked Vector Quantization
cs.CV eess.IV
This work introduces Semantically Masked VQ-GAN (SQ-GAN), a novel approach integrating generative models to optimize image compression for semantic/task-oriented communications. SQ-GAN employs off-the-shelf semantic semantic segmentation and a new specifically developed semantic-conditioned adaptive mask module (SAMM) to selectively encode semantically significant features of the images. SQ-GAN outperforms state-of-the-art image compression schemes such as JPEG2000 and BPG across multiple metrics, including perceptual quality and semantic segmentation accuracy on the post-decoding reconstructed image, at extreme low compression rates expressed in bits per pixel.
2502.09525
Robust Learning of Multi-index Models via Iterative Subspace Approximation
cs.LG cs.DS math.ST stat.ML stat.TH
We study the task of learning Multi-Index Models (MIMs) with label noise under the Gaussian distribution. A $K$-MIM is any function $f$ that only depends on a $K$-dimensional subspace. We focus on well-behaved MIMs with finite ranges that satisfy certain regularity properties. Our main contribution is a general robust learner that is qualitatively optimal in the Statistical Query (SQ) model. Our algorithm iteratively constructs better approximations to the defining subspace by computing low-degree moments conditional on the projection to the subspace computed thus far, and adding directions with relatively large empirical moments. This procedure efficiently finds a subspace $V$ so that $f(\mathbf{x})$ is close to a function of the projection of $\mathbf{x}$ onto $V$. Conversely, for functions for which these conditional moments do not help, we prove an SQ lower bound suggesting that no efficient learner exists. As applications, we provide faster robust learners for the following concept classes: * {\bf Multiclass Linear Classifiers} We give a constant-factor approximate agnostic learner with sample complexity $N = O(d) 2^{\mathrm{poly}(K/\epsilon)}$ and computational complexity $\mathrm{poly}(N ,d)$. This is the first constant-factor agnostic learner for this class whose complexity is a fixed-degree polynomial in $d$. * {\bf Intersections of Halfspaces} We give an approximate agnostic learner for this class achieving 0-1 error $K \tilde{O}(\mathrm{OPT}) + \epsilon$ with sample complexity $N=O(d^2) 2^{\mathrm{poly}(K/\epsilon)}$ and computational complexity $\mathrm{poly}(N ,d)$. This is the first agnostic learner for this class with near-linear error dependence and complexity a fixed-degree polynomial in $d$. Furthermore, we show that in the presence of random classification noise, the complexity of our algorithm scales polynomially with $1/\epsilon$.
2502.09528
SteROI-D: System Design and Mapping for Stereo Depth Inference on Regions of Interest
cs.CV cs.AR
Machine learning algorithms have enabled high quality stereo depth estimation to run on Augmented and Virtual Reality (AR/VR) devices. However, high energy consumption across the full image processing stack prevents stereo depth algorithms from running effectively on battery-limited devices. This paper introduces SteROI-D, a full stereo depth system paired with a mapping methodology. SteROI-D exploits Region-of-Interest (ROI) and temporal sparsity at the system level to save energy. SteROI-D's flexible and heterogeneous compute fabric supports diverse ROIs. Importantly, we introduce a systematic mapping methodology to effectively handle dynamic ROIs, thereby maximizing energy savings. Using these techniques, our 28nm prototype SteROI-D design achieves up to 4.35x reduction in total system energy compared to a baseline ASIC.
2502.09529
Exact Leader Estimation: A New Approach for Distributed Differentiation
eess.SY cs.MA cs.SY math.OC
A novel strategy aimed at cooperatively differentiating a signal among multiple interacting agents is introduced, where none of the agents needs to know which agent is the leader, i.e. the one producing the signal to be differentiated. Every agent communicates only a scalar variable to its neighbors; except for the leader, all agents execute the same algorithm. The proposed strategy can effectively obtain derivatives up to arbitrary $m$-th order in a finite time under the assumption that the $(m+1)$-th derivative is bounded. The strategy borrows some of its structure from the celebrated homogeneous robust exact differentiator by A. Levant, inheriting its exact differentiation capability and robustness to measurement noise. Hence, the proposed strategy can be said to perform robust exact distributed differentiation. In addition, and for the first time in the distributed leader-observer literature, sampled-data communication and bounded measurement noise are considered, and corresponding steady-state worst-case accuracy bounds are derived. The effectiveness of the proposed strategy is verified numerically for second- and fourth-order systems, i.e., for estimating derivatives of up to first and third order, respectively.
2502.09531
Data-Enabled Predictive Control for Flexible Spacecraft
eess.SY cs.SY
Spacecraft are vital to space exploration and are often equipped with lightweight, flexible appendages to meet strict weight constraints. These appendages pose significant challenges for modeling and control due to their inherent nonlinearity. Data-driven control methods have gained traction to address such challenges. This paper introduces, to the best of the authors' knowledge, the first application of the data-enabled predictive control (DeePC) framework to boundary control for flexible spacecraft. Leveraging the fundamental lemma, DeePC constructs a non-parametric model by utilizing recorded past trajectories, eliminating the need for explicit model development. The developed method also incorporates dimension reduction techniques to enhance computational efficiency. Through comprehensive numerical simulations, this study compares the proposed method with Lyapunov-based control, demonstrating superior performance and offering a thorough evaluation of data-driven control for flexible spacecraft.
2502.09532
Mind the Gap! Choice Independence in Using Multilingual LLMs for Persuasive Co-Writing Tasks in Different Languages
cs.CL cs.AI cs.HC
Recent advances in generative AI have precipitated a proliferation of novel writing assistants. These systems typically rely on multilingual large language models (LLMs), providing globalized workers the ability to revise or create diverse forms of content in different languages. However, there is substantial evidence indicating that the performance of multilingual LLMs varies between languages. Users who employ writing assistance for multiple languages are therefore susceptible to disparate output quality. Importantly, recent research has shown that people tend to generalize algorithmic errors across independent tasks, violating the behavioral axiom of choice independence. In this paper, we analyze whether user utilization of novel writing assistants in a charity advertisement writing task is affected by the AI's performance in a second language. Furthermore, we quantify the extent to which these patterns translate into the persuasiveness of generated charity advertisements, as well as the role of peoples' beliefs about LLM utilization in their donation choices. Our results provide evidence that writers who engage with an LLM-based writing assistant violate choice independence, as prior exposure to a Spanish LLM reduces subsequent utilization of an English LLM. While these patterns do not affect the aggregate persuasiveness of the generated advertisements, people's beliefs about the source of an advertisement (human versus AI) do. In particular, Spanish-speaking female participants who believed that they read an AI-generated advertisement strongly adjusted their donation behavior downwards. Furthermore, people are generally not able to adequately differentiate between human-generated and LLM-generated ads. Our work has important implications for the design, development, integration, and adoption of multilingual LLMs as assistive agents -- particularly in writing tasks.
2502.09533
Long-Term TalkingFace Generation via Motion-Prior Conditional Diffusion Model
cs.CV
Recent advances in conditional diffusion models have shown promise for generating realistic TalkingFace videos, yet challenges persist in achieving consistent head movement, synchronized facial expressions, and accurate lip synchronization over extended generations. To address these, we introduce the \textbf{M}otion-priors \textbf{C}onditional \textbf{D}iffusion \textbf{M}odel (\textbf{MCDM}), which utilizes both archived and current clip motion priors to enhance motion prediction and ensure temporal consistency. The model consists of three key elements: (1) an archived-clip motion-prior that incorporates historical frames and a reference frame to preserve identity and context; (2) a present-clip motion-prior diffusion model that captures multimodal causality for accurate predictions of head movements, lip sync, and expressions; and (3) a memory-efficient temporal attention mechanism that mitigates error accumulation by dynamically storing and updating motion features. We also release the \textbf{TalkingFace-Wild} dataset, a multilingual collection of over 200 hours of footage across 10 languages. Experimental results demonstrate the effectiveness of MCDM in maintaining identity and motion continuity for long-term TalkingFace generation. Code, models, and datasets will be publicly available.
2502.09534
Fast Tensor Completion via Approximate Richardson Iteration
cs.DS cs.LG math.ST stat.TH
We study tensor completion (TC) through the lens of low-rank tensor decomposition (TD). Many TD algorithms use fast alternating minimization methods, which solve highly structured linear regression problems at each step (e.g., for CP, Tucker, and tensor-train decompositions). However, such algebraic structure is lost in TC regression problems, making direct extensions unclear. To address this, we propose a lifting approach that approximately solves TC regression problems using structured TD regression algorithms as blackbox subroutines, enabling sublinear-time methods. We theoretically analyze the convergence rate of our approximate Richardson iteration based algorithm, and we demonstrate on real-world tensors that its running time can be 100x faster than direct methods for CP completion.
2502.09541
Vortex: Overcoming Memory Capacity Limitations in GPU-Accelerated Large-Scale Data Analytics
cs.DB cs.DC
Despite the high computational throughput of GPUs, limited memory capacity and bandwidth-limited CPU-GPU communication via PCIe links remain significant bottlenecks for accelerating large-scale data analytics workloads. This paper introduces Vortex, a GPU-accelerated framework designed for data analytics workloads that exceed GPU memory capacity. A key aspect of our framework is an optimized IO primitive that leverages all available PCIe links in multi-GPU systems for the IO demand of a single target GPU. It routes data through other GPUs to such target GPU that handles IO-intensive analytics tasks. This approach is advantageous when other GPUs are occupied with compute-bound workloads, such as popular AI applications that typically underutilize IO resources. We also introduce a novel programming model that separates GPU kernel development from IO scheduling, reducing programmer burden and enabling GPU code reuse. Additionally, we present the design of certain important query operators and discuss a late materialization technique based on GPU's zero-copy memory access. Without caching any data in GPU memory, Vortex improves the performance of the state-of-the-art GPU baseline, Proteus, by 5.7$\times$ on average and enhances price performance by 2.5$\times$ compared to a CPU-based DuckDB baseline.
2502.09553
SyntheticPop: Attacking Speaker Verification Systems With Synthetic VoicePops
cs.CR cs.LG
Voice Authentication (VA), also known as Automatic Speaker Verification (ASV), is a widely adopted authentication method, particularly in automated systems like banking services, where it serves as a secondary layer of user authentication. Despite its popularity, VA systems are vulnerable to various attacks, including replay, impersonation, and the emerging threat of deepfake audio that mimics the voice of legitimate users. To mitigate these risks, several defense mechanisms have been proposed. One such solution, Voice Pops, aims to distinguish an individual's unique phoneme pronunciations during the enrollment process. While promising, the effectiveness of VA+VoicePop against a broader range of attacks, particularly logical or adversarial attacks, remains insufficiently explored. We propose a novel attack method, which we refer to as SyntheticPop, designed to target the phoneme recognition capabilities of the VA+VoicePop system. The SyntheticPop attack involves embedding synthetic "pop" noises into spoofed audio samples, significantly degrading the model's performance. We achieve an attack success rate of over 95% while poisoning 20% of the training dataset. Our experiments demonstrate that VA+VoicePop achieves 69% accuracy under normal conditions, 37% accuracy when subjected to a baseline label flipping attack, and just 14% accuracy under our proposed SyntheticPop attack, emphasizing the effectiveness of our method.
2502.09556
Real-Time Fast Marching Tree for Mobile Robot Motion Planning in Dynamic Environments
cs.RO cs.SY eess.SY
This paper proposes the Real-Time Fast Marching Tree (RT-FMT), a real-time planning algorithm that features local and global path generation, multiple-query planning, and dynamic obstacle avoidance. During the search, RT-FMT quickly looks for the global solution and, in the meantime, generates local paths that can be used by the robot to start execution faster. In addition, our algorithm constantly rewires the tree to keep branches from forming inside the dynamic obstacles and to maintain the tree root near the robot, which allows the tree to be reused multiple times for different goals. Our algorithm is based on the planners Fast Marching Tree (FMT*) and Real-time Rapidly-Exploring Random Tree (RT-RRT*). We show via simulations that RT-FMT outperforms RT- RRT* in both execution cost and arrival time, in most cases. Moreover, we also demonstrate via simulation that it is worthwhile taking the local path before the global path is available in order to reduce arrival time, even though there is a small possibility of taking an inferior path.
2502.09560
EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents
cs.AI cs.CL cs.CV
Leveraging Multi-modal Large Language Models (MLLMs) to create embodied agents offers a promising avenue for tackling real-world tasks. While language-centric embodied agents have garnered substantial attention, MLLM-based embodied agents remain underexplored due to the lack of comprehensive evaluation frameworks. To bridge this gap, we introduce EmbodiedBench, an extensive benchmark designed to evaluate vision-driven embodied agents. EmbodiedBench features: (1) a diverse set of 1,128 testing tasks across four environments, ranging from high-level semantic tasks (e.g., household) to low-level tasks involving atomic actions (e.g., navigation and manipulation); and (2) six meticulously curated subsets evaluating essential agent capabilities like commonsense reasoning, complex instruction understanding, spatial awareness, visual perception, and long-term planning. Through extensive experiments, we evaluated 13 leading proprietary and open-source MLLMs within EmbodiedBench. Our findings reveal that: MLLMs excel at high-level tasks but struggle with low-level manipulation, with the best model, GPT-4o, scoring only 28.9% on average. EmbodiedBench provides a multifaceted standardized evaluation platform that not only highlights existing challenges but also offers valuable insights to advance MLLM-based embodied agents. Our code is available at https://embodiedbench.github.io.
2502.09561
Enhancing Traffic Safety Analysis with Digital Twin Technology: Integrating Vehicle Dynamics and Environmental Factors into Microscopic Traffic Simulation
eess.SY cs.SY math.OC
Traffic safety is a critical concern in transportation engineering and urban planning. Traditional traffic safety analysis requires trained observers to collect data in the field, which is time-consuming, labor-intensive, and sometimes inaccurate. In recent years, microscopic traffic simulation, which simulates individual vehicles' movements within a transportation network, have been utilized to study traffic safety. However, microscopic traffic simulation only focuses on traffic-related factors, such as traffic volume, traffic signals, and lane configurations, neglecting vehicle dynamics and environment-related factors like weather and lighting conditions, which can significantly impact traffic safety. In light of this, this paper explores the application of digital twin technology in traffic safety analysis, integrating vehicle simulators, which consider vehicle dynamics and environmental factors, and microscopic traffic simulators, which simulate the operations of traffic flow, for enhanced safety evaluations. Various scenarios, including different weather conditions and visibility levels, are simulated using a digital twin of a road segment in Tuscaloosa, Alabama. The simulations employ Surrogate Safety Measures (SSMs) like Time to Collision (TTC) and Deceleration Rate to Avoid a Crash (DRAC) to assess safety under varying conditions. The results demonstrate that traffic digital twin can identify potential safety issues that traditional microscopic simulation cannot, providing insights for improving traffic control strategies and transportation infrastructure to enhance traffic safety.
2502.09563
Self-Calibrating Gaussian Splatting for Large Field of View Reconstruction
cs.CV cs.GR
In this paper, we present a self-calibrating framework that jointly optimizes camera parameters, lens distortion and 3D Gaussian representations, enabling accurate and efficient scene reconstruction. In particular, our technique enables high-quality scene reconstruction from Large field-of-view (FOV) imagery taken with wide-angle lenses, allowing the scene to be modeled from a smaller number of images. Our approach introduces a novel method for modeling complex lens distortions using a hybrid network that combines invertible residual networks with explicit grids. This design effectively regularizes the optimization process, achieving greater accuracy than conventional camera models. Additionally, we propose a cubemap-based resampling strategy to support large FOV images without sacrificing resolution or introducing distortion artifacts. Our method is compatible with the fast rasterization of Gaussian Splatting, adaptable to a wide variety of camera lens distortion, and demonstrates state-of-the-art performance on both synthetic and real-world datasets.
2502.09564
Diffusing DeBias: a Recipe for Turning a Bug into a Feature
cs.LG cs.CV
Deep learning model effectiveness in classification tasks is often challenged by the quality and quantity of training data which, whenever containing strong spurious correlations between specific attributes and target labels, can result in unrecoverable biases in model predictions. Tackling these biases is crucial in improving model generalization and trust, especially in real-world scenarios. This paper presents Diffusing DeBias (DDB), a novel approach acting as a plug-in for common methods in model debiasing while exploiting the inherent bias-learning tendency of diffusion models. Our approach leverages conditional diffusion models to generate synthetic bias-aligned images, used to train a bias amplifier model, to be further employed as an auxiliary method in different unsupervised debiasing approaches. Our proposed method, which also tackles the common issue of training set memorization typical of this type of techniques, beats current state-of-the-art in multiple benchmark datasets by significant margins, demonstrating its potential as a versatile and effective tool for tackling dataset bias in deep learning applications.
2502.09565
MDCrow: Automating Molecular Dynamics Workflows with Large Language Models
cs.AI physics.chem-ph
Molecular dynamics (MD) simulations are essential for understanding biomolecular systems but remain challenging to automate. Recent advances in large language models (LLM) have demonstrated success in automating complex scientific tasks using LLM-based agents. In this paper, we introduce MDCrow, an agentic LLM assistant capable of automating MD workflows. MDCrow uses chain-of-thought over 40 expert-designed tools for handling and processing files, setting up simulations, analyzing the simulation outputs, and retrieving relevant information from literature and databases. We assess MDCrow's performance across 25 tasks of varying required subtasks and difficulty, and we evaluate the agent's robustness to both difficulty and prompt style. \texttt{gpt-4o} is able to complete complex tasks with low variance, followed closely by \texttt{llama3-405b}, a compelling open-source model. While prompt style does not influence the best models' performance, it has significant effects on smaller models.
2502.09566
Zero-shot generation of synthetic neurosurgical data with large language models
cs.CL cs.LG
Clinical data is fundamental to advance neurosurgical research, but access is often constrained by data availability, small sample sizes, privacy regulations, and resource-intensive preprocessing and de-identification procedures. Synthetic data offers a potential solution to challenges associated with accessing and using real-world data (RWD). This study aims to evaluate the capability of zero-shot generation of synthetic neurosurgical data with a large language model (LLM), GPT-4o, by benchmarking with the conditional tabular generative adversarial network (CTGAN). Synthetic datasets were compared to real-world neurosurgical data to assess fidelity (means, proportions, distributions, and bivariate correlations), utility (ML classifier performance on RWD), and privacy (duplication of records from RWD). The GPT-4o-generated datasets matched or exceeded CTGAN performance, despite no fine-tuning or access to RWD for pre-training. Datasets demonstrated high univariate and bivariate fidelity to RWD without directly exposing any real patient records, even at amplified sample size. Training an ML classifier on GPT-4o-generated data and testing on RWD for a binary prediction task showed an F1 score (0.706) with comparable performance to training on the CTGAN data (0.705) for predicting postoperative functional status deterioration. GPT-4o demonstrated a promising ability to generate high-fidelity synthetic neurosurgical data. These findings also indicate that data synthesized with GPT-4o can effectively augment clinical data with small sample sizes, and train ML models for prediction of neurosurgical outcomes. Further investigation is necessary to improve the preservation of distributional characteristics and boost classifier performance.
2502.09567
MorphNLI: A Stepwise Approach to Natural Language Inference Using Text Morphing
cs.CL cs.AI
We introduce MorphNLI, a modular step-by-step approach to natural language inference (NLI). When classifying the premise-hypothesis pairs into {entailment, contradiction, neutral}, we use a language model to generate the necessary edits to incrementally transform (i.e., morph) the premise into the hypothesis. Then, using an off-the-shelf NLI model we track how the entailment progresses with these atomic changes, aggregating these intermediate labels into a final output. We demonstrate the advantages of our proposed method particularly in realistic cross-domain settings, where our method always outperforms strong baselines with improvements up to 12.6% (relative). Further, our proposed approach is explainable as the atomic edits can be used to understand the overall NLI label.
2502.09570
Enhancing the Utility of Higher-Order Information in Relational Learning
cs.LG stat.ML
Higher-order information is crucial for relational learning in many domains where relationships extend beyond pairwise interactions. Hypergraphs provide a natural framework for modeling such relationships, which has motivated recent extensions of graph neural network architectures to hypergraphs. However, comparisons between hypergraph architectures and standard graph-level models remain limited. In this work, we systematically evaluate a selection of hypergraph-level and graph-level architectures, to determine their effectiveness in leveraging higher-order information in relational learning. Our results show that graph-level architectures applied to hypergraph expansions often outperform hypergraph-level ones, even on inputs that are naturally parametrized as hypergraphs. As an alternative approach for leveraging higher-order information, we propose hypergraph-level encodings based on classical hypergraph characteristics. While these encodings do not significantly improve hypergraph architectures, they yield substantial performance gains when combined with graph-level models. Our theoretical analysis shows that hypergraph-level encodings provably increase the representational power of message-passing graph neural networks beyond that of their graph-level counterparts.
2502.09571
DiffMS: Diffusion Generation of Molecules Conditioned on Mass Spectra
cs.LG q-bio.QM
Mass spectrometry plays a fundamental role in elucidating the structures of unknown molecules and subsequent scientific discoveries. One formulation of the structure elucidation task is the conditional $\textit{de novo}$ generation of molecular structure given a mass spectrum. Toward a more accurate and efficient scientific discovery pipeline for small molecules, we present DiffMS, a formula-restricted encoder-decoder generative network that achieves state-of-the-art performance on this task. The encoder utilizes a transformer architecture and models mass spectra domain knowledge such as peak formulae and neutral losses, and the decoder is a discrete graph diffusion model restricted by the heavy-atom composition of a known chemical formula. To develop a robust decoder that bridges latent embeddings and molecular structures, we pretrain the diffusion decoder with fingerprint-structure pairs, which are available in virtually infinite quantities, compared to structure-spectrum pairs that number in the tens of thousands. Extensive experiments on established benchmarks show that DiffMS outperforms existing models on $\textit{de novo}$ molecule generation. We provide several ablations to demonstrate the effectiveness of our diffusion and pretraining approaches and show consistent performance scaling with increasing pretraining dataset size. DiffMS code is publicly available at https://github.com/coleygroup/DiffMS.
2502.09573
Optimizing GPT for Video Understanding: Zero-Shot Performance and Prompt Engineering
cs.CV cs.CL cs.LG
In this study, we tackle industry challenges in video content classification by exploring and optimizing GPT-based models for zero-shot classification across seven critical categories of video quality. We contribute a novel approach to improving GPT's performance through prompt optimization and policy refinement, demonstrating that simplifying complex policies significantly reduces false negatives. Additionally, we introduce a new decomposition-aggregation-based prompt engineering technique, which outperforms traditional single-prompt methods. These experiments, conducted on real industry problems, show that thoughtful prompt design can substantially enhance GPT's performance without additional finetuning, offering an effective and scalable solution for improving video classification systems across various domains in industry.
2502.09583
Learning to Coordinate with Experts
cs.LG stat.ML
When deployed in dynamic environments, AI agents will inevitably encounter challenges that exceed their individual capabilities. Leveraging assistance from expert agents-whether human or AI-can significantly enhance safety and performance in such situations. However, querying experts is often costly, necessitating the development of agents that can efficiently request and utilize expert guidance. In this paper, we introduce a fundamental coordination problem called Learning to Yield and Request Control (YRC), where the objective is to learn a strategy that determines when to act autonomously and when to seek expert assistance. We consider a challenging practical setting in which an agent does not interact with experts during training but must adapt to novel environmental changes and expert interventions at test time. To facilitate empirical research, we introduce YRC-Bench, an open-source benchmark featuring diverse domains. YRC-Bench provides a standardized Gym-like API, simulated experts, evaluation pipeline, and implementation of competitive baselines. Towards tackling the YRC problem, we propose a novel validation approach and investigate the performance of various learning methods across diverse environments, yielding insights that can guide future research.
2502.09587
Rolling Ahead Diffusion for Traffic Scene Simulation
cs.LG cs.RO
Realistic driving simulation requires that NPCs not only mimic natural driving behaviors but also react to the behavior of other simulated agents. Recent developments in diffusion-based scenario generation focus on creating diverse and realistic traffic scenarios by jointly modelling the motion of all the agents in the scene. However, these traffic scenarios do not react when the motion of agents deviates from their modelled trajectories. For example, the ego-agent can be controlled by a stand along motion planner. To produce reactive scenarios with joint scenario models, the model must regenerate the scenario at each timestep based on new observations in a Model Predictive Control (MPC) fashion. Although reactive, this method is time-consuming, as one complete possible future for all NPCs is generated per simulation step. Alternatively, one can utilize an autoregressive model (AR) to predict only the immediate next-step future for all NPCs. Although faster, this method lacks the capability for advanced planning. We present a rolling diffusion based traffic scene generation model which mixes the benefits of both methods by predicting the next step future and simultaneously predicting partially noised further future steps at the same time. We show that such model is efficient compared to diffusion model based AR, achieving a beneficial compromise between reactivity and computational efficiency.
2502.09589
Logical forms complement probability in understanding language model (and human) performance
cs.CL cs.LO
With the increasing interest in using large language models (LLMs) for planning in natural language, understanding their behaviors becomes an important research question. This work conducts a systematic investigation of LLMs' ability to perform logical reasoning in natural language. We introduce a controlled dataset of hypothetical and disjunctive syllogisms in propositional and modal logic and use it as the testbed for understanding LLM performance. Our results lead to novel insights in predicting LLM behaviors: in addition to the probability of input (Gonen et al., 2023; McCoy et al., 2024), logical forms should be considered as important factors. In addition, we show similarities and discrepancies between the logical reasoning performances of humans and LLMs by collecting and comparing behavioral data from both.