id
stringlengths
9
16
title
stringlengths
4
278
categories
stringlengths
5
104
abstract
stringlengths
6
4.09k
2501.13734
Sample complexity of data-driven tuning of model hyperparameters in neural networks with structured parameter-dependent dual function
cs.LG
Modern machine learning algorithms, especially deep learning based techniques, typically involve careful hyperparameter tuning to achieve the best performance. Despite the surge of intense interest in practical techniques like Bayesian optimization and random search based approaches to automating this laborious and compute intensive task, the fundamental learning theoretic complexity of tuning hyperparameters for deep neural networks is poorly understood. Inspired by this glaring gap, we initiate the formal study of hyperparameter tuning complexity in deep learning through a recently introduced data driven setting. We assume that we have a series of deep learning tasks, and we have to tune hyperparameters to do well on average over the distribution of tasks. A major difficulty is that the utility function as a function of the hyperparameter is very volatile and furthermore, it is given implicitly by an optimization problem over the model parameters. To tackle this challenge, we introduce a new technique to characterize the discontinuities and oscillations of the utility function on any fixed problem instance as we vary the hyperparameter; our analysis relies on subtle concepts including tools from differential/algebraic geometry and constrained optimization. This can be used to show that the learning theoretic complexity of the corresponding family of utility functions is bounded. We instantiate our results and provide sample complexity bounds for concrete applications tuning a hyperparameter that interpolates neural activation functions and setting the kernel parameter in graph neural networks.
2501.13735
A Study of the Plausibility of Attention between RNN Encoders in Natural Language Inference
cs.CL
Attention maps in neural models for NLP are appealing to explain the decision made by a model, hopefully emphasizing words that justify the decision. While many empirical studies hint that attention maps can provide such justification from the analysis of sound examples, only a few assess the plausibility of explanations based on attention maps, i.e., the usefulness of attention maps for humans to understand the decision. These studies furthermore focus on text classification. In this paper, we report on a preliminary assessment of attention maps in a sentence comparison task, namely natural language inference. We compare the cross-attention weights between two RNN encoders with human-based and heuristic-based annotations on the eSNLI corpus. We show that the heuristic reasonably correlates with human annotations and can thus facilitate evaluation of plausible explanations in sentence comparison tasks. Raw attention weights however remain only loosely related to a plausible explanation.
2501.13736
Discrete Layered Entropy, Conditional Compression and a Tighter Strong Functional Representation Lemma
cs.IT math.IT
We study a quantity called discrete layered entropy, which approximates the Shannon entropy within a logarithmic gap. Compared to the Shannon entropy, the discrete layered entropy is piecewise linear, approximates the expected length of the optimal one-to-one non-prefix-free encoding, and satisfies an elegant conditioning property. These properties make it useful for approximating the Shannon entropy in linear programming, studying the optimal length of conditional encoding, and bounding the entropy of monotonic mixture distributions. In particular, it can give a bound for the strong functional representation lemma that improves upon the best bound (as long as the mutual information is at least 2).
2501.13743
GPT-HTree: A Decision Tree Framework Integrating Hierarchical Clustering and Large Language Models for Explainable Classification
cs.LG
This paper introduces GPT-HTree, a framework combining hierarchical clustering, decision trees, and large language models (LLMs) to address this challenge. By leveraging hierarchical clustering to segment individuals based on salient features, resampling techniques to balance class distributions, and decision trees to tailor classification paths within each cluster, GPT-HTree ensures both accuracy and interpretability. LLMs enhance the framework by generating human-readable cluster descriptions, bridging quantitative analysis with actionable insights.
2501.13744
Centralized Versus Distributed Routing for Large-Scale Satellite Networks
cs.NI cs.SY eess.SY
An important choice in the design of satellite networks is whether the routing decisions are made in a distributed manner onboard the satellite, or centrally on a ground-based controller. We study the tradeoff between centralized and distributed routing in large-scale satellite networks. In particular, we consider a centralized routing scheme that has access to global but delayed network state information and a distributed routing scheme that has access to local but real-time network state information. For both routing schemes, we analyze the throughput and delay performance of shortest-path algorithms in networks with and without buffers onboard the satellites. We show that distributed routing outperforms centralized routing when the rate of changes in the network link state is comparable to the inherent propagation and transmission delays. In particular, we show that in highly dynamic networks without buffers, the distributed scheme achieves higher throughput than a centralized scheme. In networks with buffers, the distributed scheme achieves lower delays with the same throughput.
2501.13746
EICopilot: Search and Explore Enterprise Information over Large-scale Knowledge Graphs with LLM-driven Agents
cs.IR cs.AI
The paper introduces EICopilot, an novel agent-based solution enhancing search and exploration of enterprise registration data within extensive online knowledge graphs like those detailing legal entities, registered capital, and major shareholders. Traditional methods necessitate text-based queries and manual subgraph explorations, often resulting in time-consuming processes. EICopilot, deployed as a chatbot via Baidu Enterprise Search, improves this landscape by utilizing Large Language Models (LLMs) to interpret natural language queries. This solution automatically generates and executes Gremlin scripts, providing efficient summaries of complex enterprise relationships. Distinct feature a data pre-processing pipeline that compiles and annotates representative queries into a vector database of examples for In-context learning (ICL), a comprehensive reasoning pipeline combining Chain-of-Thought with ICL to enhance Gremlin script generation for knowledge graph search and exploration, and a novel query masking strategy that improves intent recognition for heightened script accuracy. Empirical evaluations demonstrate the superior performance of EICopilot, including speed and accuracy, over baseline methods, with the \emph{Full Mask} variant achieving a syntax error rate reduction to as low as 10.00% and an execution correctness of up to 82.14%. These components collectively contribute to superior querying capabilities and summarization of intricate datasets, positioning EICopilot as a groundbreaking tool in the exploration and exploitation of large-scale knowledge graphs for enterprise information search.
2501.13748
Exact Soft Analytical Side-Channel Attacks using Tractable Circuits
cs.LG cs.CR
Detecting weaknesses in cryptographic algorithms is of utmost importance for designing secure information systems. The state-of-the-art soft analytical side-channel attack (SASCA) uses physical leakage information to make probabilistic predictions about intermediate computations and combines these "guesses" with the known algorithmic logic to compute the posterior distribution over the key. This attack is commonly performed via loopy belief propagation, which, however, lacks guarantees in terms of convergence and inference quality. In this paper, we develop a fast and exact inference method for SASCA, denoted as ExSASCA, by leveraging knowledge compilation and tractable probabilistic circuits. When attacking the Advanced Encryption Standard (AES), the most widely used encryption algorithm to date, ExSASCA outperforms SASCA by more than 31% top-1 success rate absolute. By leveraging sparse belief messages, this performance is achieved with little more computational cost than SASCA, and about 3 orders of magnitude less than exact inference via exhaustive enumeration. Even with dense belief messages, ExSASCA still uses 6 times less computations than exhaustive inference.
2501.13751
On Disentangled Training for Nonlinear Transform in Learned Image Compression
eess.IV cs.CV
Learned image compression (LIC) has demonstrated superior rate-distortion (R-D) performance compared to traditional codecs, but is challenged by training inefficiency that could incur more than two weeks to train a state-of-the-art model from scratch. Existing LIC methods overlook the slow convergence caused by compacting energy in learning nonlinear transforms. In this paper, we first reveal that such energy compaction consists of two components, i.e., feature decorrelation and uneven energy modulation. On such basis, we propose a linear auxiliary transform (AuxT) to disentangle energy compaction in training nonlinear transforms. The proposed AuxT obtains coarse approximation to achieve efficient energy compaction such that distribution fitting with the nonlinear transforms can be simplified to fine details. We then develop wavelet-based linear shortcuts (WLSs) for AuxT that leverages wavelet-based downsampling and orthogonal linear projection for feature decorrelation and subband-aware scaling for
2501.13756
Solving the long-tailed distribution problem by exploiting the synergies and balance of different techniques
cs.CV cs.AI cs.LG
In real-world data, long-tailed data distribution is common, making it challenging for models trained on empirical risk minimisation to learn and classify tail classes effectively. While many studies have sought to improve long tail recognition by altering the data distribution in the feature space and adjusting model decision boundaries, research on the synergy and corrective approach among various methods is limited. Our study delves into three long-tail recognition techniques: Supervised Contrastive Learning (SCL), Rare-Class Sample Generator (RSG), and Label-Distribution-Aware Margin Loss (LDAM). SCL enhances intra-class clusters based on feature similarity and promotes clear inter-class separability but tends to favour dominant classes only. When RSG is integrated into the model, we observed that the intra-class features further cluster towards the class centre, which demonstrates a synergistic effect together with SCL's principle of enhancing intra-class clustering. RSG generates new tail features and compensates for the tail feature space squeezed by SCL. Similarly, LDAM is known to introduce a larger margin specifically for tail classes; we demonstrate that LDAM further bolsters the model's performance on tail classes when combined with the more explicit decision boundaries achieved by SCL and RSG. Furthermore, SCL can compensate for the dominant class accuracy sacrificed by RSG and LDAM. Our research emphasises the synergy and balance among the three techniques, with each amplifying the strengths of the others and mitigating their shortcomings. Our experiment on long-tailed distribution datasets, using an end-to-end architecture, yields competitive results by enhancing tail class accuracy without compromising dominant class performance, achieving a balanced improvement across all classes.
2501.13758
2-Tier SimCSE: Elevating BERT for Robust Sentence Embeddings
cs.CL cs.AI cs.LG
Effective sentence embeddings that capture semantic nuances and generalize well across diverse contexts are crucial for natural language processing tasks. We address this challenge by applying SimCSE (Simple Contrastive Learning of Sentence Embeddings) using contrastive learning to fine-tune the minBERT model for sentiment analysis, semantic textual similarity (STS), and paraphrase detection. Our contributions include experimenting with three different dropout techniques, namely standard dropout, curriculum dropout, and adaptive dropout, to tackle overfitting, proposing a novel 2-Tier SimCSE Fine-tuning Model that combines both unsupervised and supervised SimCSE on STS task, and exploring transfer learning potential for Paraphrase and SST tasks. Our findings demonstrate the effectiveness of SimCSE, with the 2-Tier model achieving superior performance on the STS task, with an average test score of 0.742 across all three downstream tasks. The results of error analysis reveals challenges in handling complex sentiments and reliance on lexical overlap for paraphrase detection, highlighting areas for future research. The ablation study revealed that removing Adaptive Dropout in the Single-Task Unsupervised SimCSE Model led to improved performance on the STS task, indicating overfitting due to added parameters. Transfer learning from SimCSE models on Paraphrase and SST tasks did not enhance performance, suggesting limited transferability of knowledge from the STS task.
2501.13762
On Deciding the Data Complexity of Answering Linear Monadic Datalog Queries with LTL Operators(Extended Version)
cs.AI cs.CC cs.LO
Our concern is the data complexity of answering linear monadic datalog queries whose atoms in the rule bodies can be prefixed by operators of linear temporal logic LTL. We first observe that, for data complexity, answering any connected query with operators $\bigcirc/\bigcirc^-$ (at the next/previous moment) is either in AC0, or in $ACC0\!\setminus\!AC0$, or $NC^1$-complete, or LogSpace-hard and in NLogSpace. Then we show that the problem of deciding LogSpace-hardness of answering such queries is PSpace-complete, while checking membership in the classes AC0 and ACC0 as well as $NC^1$-completeness can be done in ExpSpace. Finally, we prove that membership in AC0 or in ACC0, $NC^1$-completeness, and LogSpace-hardness are undecidable for queries with operators $\Diamond_f/\Diamond_p$ (sometime in the future/past) provided that $NC^1 \ne NLogSpace$, and $LogSpace \ne NLogSpace$.
2501.13763
Integrating Causality with Neurochaos Learning: Proposed Approach and Research Agenda
cs.LG cs.AI
Deep learning implemented via neural networks, has revolutionized machine learning by providing methods for complex tasks such as object detection/classification and prediction. However, architectures based on deep neural networks have started to yield diminishing returns, primarily due to their statistical nature and inability to capture causal structure in the training data. Another issue with deep learning is its high energy consumption, which is not that desirable from a sustainability perspective. Therefore, alternative approaches are being considered to address these issues, both of which are inspired by the functioning of the human brain. One approach is causal learning, which takes into account causality among the items in the dataset on which the neural network is trained. It is expected that this will help minimize the spurious correlations that are prevalent in the learned representations of deep neural networks. The other approach is Neurochaos Learning, a recent development, which draws its inspiration from the nonlinear chaotic firing intrinsic to neurons in biological neural networks (brain/central nervous system). Both approaches have shown improved results over just deep learning alone. To that end, in this position paper, we investigate how causal and neurochaos learning approaches can be integrated together to produce better results, especially in domains that contain linked data. We propose an approach for this integration to enhance classification, prediction and reinforcement learning. We also propose a set of research questions that need to be investigated in order to make this integration a reality.
2501.13766
UGMathBench: A Diverse and Dynamic Benchmark for Undergraduate-Level Mathematical Reasoning with Large Language Models
cs.CL cs.AI
Large Language Models (LLMs) have made significant strides in mathematical reasoning, underscoring the need for a comprehensive and fair evaluation of their capabilities. However, existing benchmarks often fall short, either lacking extensive coverage of undergraduate-level mathematical problems or probably suffering from test-set contamination. To address these issues, we introduce UGMathBench, a diverse and dynamic benchmark specifically designed for evaluating undergraduate-level mathematical reasoning with LLMs. UGMathBench comprises 5,062 problems across 16 subjects and 111 topics, featuring 10 distinct answer types. Each problem includes three randomized versions, with additional versions planned for release as leading open-source LLMs become saturated in UGMathBench. Furthermore, we propose two key metrics: effective accuracy (EAcc), which measures the percentage of correctly solved problems across all three versions, and reasoning gap ($\Delta$), which assesses reasoning robustness by calculating the difference between the average accuracy across all versions and EAcc. Our extensive evaluation of 23 leading LLMs reveals that the highest EAcc achieved is 56.3\% by OpenAI-o1-mini, with large $\Delta$ values observed across different models. This highlights the need for future research aimed at developing "large reasoning models" with high EAcc and $\Delta = 0$. We anticipate that the release of UGMathBench, along with its detailed evaluation codes, will serve as a valuable resource to advance the development of LLMs in solving mathematical problems.
2501.13767
An Efficient Diffusion-based Non-Autoregressive Solver for Traveling Salesman Problem
cs.LG
Recent advances in neural models have shown considerable promise in solving Traveling Salesman Problems (TSPs) without relying on much hand-crafted engineering. However, while non-autoregressive (NAR) approaches benefit from faster inference through parallelism, they typically deliver solutions of inferior quality compared to autoregressive ones. To enhance the solution quality while maintaining fast inference, we propose DEITSP, a diffusion model with efficient iterations tailored for TSP that operates in a NAR manner. Firstly, we introduce a one-step diffusion model that integrates the controlled discrete noise addition process with self-consistency enhancement, enabling optimal solution prediction through simultaneous denoising of multiple solutions. Secondly, we design a dual-modality graph transformer to bolster the extraction and fusion of features from node and edge modalities, while further accelerating the inference with fewer layers. Thirdly, we develop an efficient iterative strategy that alternates between adding and removing noise to improve exploration compared to previous diffusion methods. Additionally, we devise a scheduling framework to progressively refine the solution space by adjusting noise levels, facilitating a smooth search for optimal solutions. Extensive experiments on real-world and large-scale TSP instances demonstrate that DEITSP performs favorably against existing neural approaches in terms of solution quality, inference latency, and generalization ability. Our code is available at $\href{https://github.com/DEITSP/DEITSP}{https://github.com/DEITSP/DEITSP}$.
2501.13772
Tune In, Act Up: Exploring the Impact of Audio Modality-Specific Edits on Large Audio Language Models in Jailbreak
cs.SD cs.AI cs.LG cs.MM eess.AS
Large Language Models (LLMs) demonstrate remarkable zero-shot performance across various natural language processing tasks. The integration of multimodal encoders extends their capabilities, enabling the development of Multimodal Large Language Models that process vision, audio, and text. However, these capabilities also raise significant security concerns, as these models can be manipulated to generate harmful or inappropriate content through jailbreak. While extensive research explores the impact of modality-specific input edits on text-based LLMs and Large Vision-Language Models in jailbreak, the effects of audio-specific edits on Large Audio-Language Models (LALMs) remain underexplored. Hence, this paper addresses this gap by investigating how audio-specific edits influence LALMs inference regarding jailbreak. We introduce the Audio Editing Toolbox (AET), which enables audio-modality edits such as tone adjustment, word emphasis, and noise injection, and the Edited Audio Datasets (EADs), a comprehensive audio jailbreak benchmark. We also conduct extensive evaluations of state-of-the-art LALMs to assess their robustness under different audio edits. This work lays the groundwork for future explorations on audio-modality interactions in LALMs security.
2501.13773
Do Large Language Models Truly Understand Geometric Structures?
cs.CL
Geometric ability is a significant challenge for large language models (LLMs) due to the need for advanced spatial comprehension and abstract thinking. Existing datasets primarily evaluate LLMs on their final answers, but they cannot truly measure their true understanding of geometric structures, as LLMs can arrive at correct answers by coincidence. To fill this gap, we introduce the GeomRel dataset, designed to evaluate LLMs' understanding of geometric structures by isolating the core step of geometric relationship identification in problem-solving. Using this benchmark, we conduct thorough evaluations of diverse LLMs and identify key limitations in understanding geometric structures. We further propose the Geometry Chain-of-Thought (GeoCoT) method, which enhances LLMs' ability to identify geometric relationships, resulting in significant performance improvements.
2501.13776
Crossfire: An Elastic Defense Framework for Graph Neural Networks Under Bit Flip Attacks
cs.LG
Bit Flip Attacks (BFAs) are a well-established class of adversarial attacks, originally developed for Convolutional Neural Networks within the computer vision domain. Most recently, these attacks have been extended to target Graph Neural Networks (GNNs), revealing significant vulnerabilities. This new development naturally raises questions about the best strategies to defend GNNs against BFAs, a challenge for which no solutions currently exist. Given the applications of GNNs in critical fields, any defense mechanism must not only maintain network performance, but also verifiably restore the network to its pre-attack state. Verifiably restoring the network to its pre-attack state also eliminates the need for costly evaluations on test data to ensure network quality. We offer first insights into the effectiveness of existing honeypot- and hashing-based defenses against BFAs adapted from the computer vision domain to GNNs, and characterize the shortcomings of these approaches. To overcome their limitations, we propose Crossfire, a hybrid approach that exploits weight sparsity and combines hashing and honeypots with bit-level correction of out-of-distribution weight elements to restore network integrity. Crossfire is retraining-free and does not require labeled data. Averaged over 2,160 experiments on six benchmark datasets, Crossfire offers a 21.8% higher probability than its competitors of reconstructing a GNN attacked by a BFA to its pre-attack state. These experiments cover up to 55 bit flips from various attacks. Moreover, it improves post-repair prediction quality by 10.85%. Computational and storage overheads are negligible compared to the inherent complexity of even the simplest GNNs.
2501.13778
Explainable XR: Understanding User Behaviors of XR Environments using LLM-assisted Analytics Framework
cs.HC cs.CL
We present Explainable XR, an end-to-end framework for analyzing user behavior in diverse eXtended Reality (XR) environments by leveraging Large Language Models (LLMs) for data interpretation assistance. Existing XR user analytics frameworks face challenges in handling cross-virtuality - AR, VR, MR - transitions, multi-user collaborative application scenarios, and the complexity of multimodal data. Explainable XR addresses these challenges by providing a virtuality-agnostic solution for the collection, analysis, and visualization of immersive sessions. We propose three main components in our framework: (1) A novel user data recording schema, called User Action Descriptor (UAD), that can capture the users' multimodal actions, along with their intents and the contexts; (2) a platform-agnostic XR session recorder, and (3) a visual analytics interface that offers LLM-assisted insights tailored to the analysts' perspectives, facilitating the exploration and analysis of the recorded XR session data. We demonstrate the versatility of Explainable XR by demonstrating five use-case scenarios, in both individual and collaborative XR applications across virtualities. Our technical evaluation and user studies show that Explainable XR provides a highly usable analytics solution for understanding user actions and delivering multifaceted, actionable insights into user behaviors in immersive environments.
2501.13779
Not Every AI Problem is a Data Problem: We Should Be Intentional About Data Scaling
cs.LG cs.AI
While Large Language Models require more and more data to train and scale, rather than looking for any data to acquire, we should consider what types of tasks are more likely to benefit from data scaling. We should be intentional in our data acquisition. We argue that the topology of data itself informs which tasks to prioritize in data scaling, and shapes the development of the next generation of compute paradigms for tasks where data scaling is inefficient, or even insufficient.
2501.13780
Matrix Completion in Group Testing: Bounds and Simulations
cs.IT cs.LG math.IT
The main goal of group testing is to identify a small number of defective items in a large population of items. A test on a subset of items is positive if the subset contains at least one defective item and negative otherwise. In non-adaptive design, all tests can be tested simultaneously and represented by a measurement matrix in which a row and a column represent a test and an item, respectively. An entry in row $i$ and column $j$ is 1 if item $j$ belongs to the test $i$ and is 0 otherwise. Given an unknown set of defective items, the objective is to design a measurement matrix such that, by observing its corresponding outcome vector, the defective items can be recovered efficiently. The basic trait of this approach is that the measurement matrix has remained unchanged throughout the course of generating the outcome vector and recovering defective items. In this paper, we study the case in which some entries in the measurement matrix are erased, called \emph{the missing measurement matrix}, before the recovery phase of the defective items, and our objective is to fully recover the measurement matrix from the missing measurement matrix. In particular, we show that some specific rows with erased entries provide information aiding the recovery while others do not. Given measurement matrices and erased entries follow the Bernoulli distribution, we show that before the erasing event happens, sampling sufficient sets of defective items and their corresponding outcome vectors can help us recover the measurement matrix from the missing measurement matrix.
2501.13782
Defending against Adversarial Malware Attacks on ML-based Android Malware Detection Systems
cs.CR cs.AI cs.LG cs.SE
Android malware presents a persistent threat to users' privacy and data integrity. To combat this, researchers have proposed machine learning-based (ML-based) Android malware detection (AMD) systems. However, adversarial Android malware attacks compromise the detection integrity of the ML-based AMD systems, raising significant concerns. Existing defenses against adversarial Android malware provide protections against feature space attacks which generate adversarial feature vectors only, leaving protection against realistic threats from problem space attacks which generate real adversarial malware an open problem. In this paper, we address this gap by proposing ADD, a practical adversarial Android malware defense framework designed as a plug-in to enhance the adversarial robustness of the ML-based AMD systems against problem space attacks. Our extensive evaluation across various ML-based AMD systems demonstrates that ADD is effective against state-of-the-art problem space adversarial Android malware attacks. Additionally, ADD shows the defense effectiveness in enhancing the adversarial robustness of real-world antivirus solutions.
2501.13784
Rate-Distortion Region for Distributed Indirect Source Coding with Decoder Side Information
cs.IT math.IT
This paper studies a variant of the rate-distortion problem motivated by task-oriented semantic communication and distributed learning systems, where $M$ correlated sources are independently encoded for a central decoder. The decoder has access to correlated side information in addition to the messages received from the encoders and aims to recover a latent random variable under a given distortion constraint, rather than recovering the sources themselves. We characterize the exact rate-distortion function for the case where the sources are conditionally independent given the side information. Furthermore, we develop a distributed Blahut-Arimoto (BA) algorithm to numerically compute the rate-distortion function. Numerical examples are provided to demonstrate the effectiveness of the proposed approach in calculating the rate-distortion region.
2501.13786
Fast Iterative and Task-Specific Imputation with Online Learning
cs.LG
Missing feature values are a significant hurdle for downstream machine-learning tasks such as classification and regression. However, they are pervasive in multiple real-life use cases, for instance, in drug discovery research. Moreover, imputation methods might be time-consuming and offer few guarantees on the imputation quality, especially for not-missing-at-random mechanisms. We propose an imputation approach named F3I based on the iterative improvement of a K-nearest neighbor imputation that learns the weights for each neighbor of a data point, optimizing for the most likely distribution of points over data points. This algorithm can also be jointly trained with a downstream task on the imputed values. We provide a theoretical analysis of the imputation quality by F3I for several types of missing mechanisms. We also demonstrate the performance of F3I on both synthetic data sets and real-life drug repurposing and handwritten-digit recognition data.
2501.13787
Parameter-Efficient Fine-Tuning for Foundation Models
cs.CL cs.AI cs.LG
This survey delves into the realm of Parameter-Efficient Fine-Tuning (PEFT) within the context of Foundation Models (FMs). PEFT, a cost-effective fine-tuning technique, minimizes parameters and computational complexity while striving for optimal downstream task performance. FMs, like ChatGPT, DALL-E, and LLaVA specialize in language understanding, generative tasks, and multimodal tasks, trained on diverse datasets spanning text, images, and videos. The diversity of FMs guides various adaptation strategies for PEFT. Therefore, this survey aims to provide a comprehensive overview of PEFT techniques applied to diverse FMs and address critical gaps in understanding the techniques, trends, and applications. We start by providing a detailed development of FMs and PEFT. Subsequently, we systematically review the key categories and core mechanisms of PEFT across diverse FMs to offer a comprehensive understanding of trends. We also explore the most recent applications across various FMs to demonstrate the versatility of PEFT, shedding light on the integration of systematic PEFT methods with a range of FMs. Furthermore, we identify potential research and development directions for improving PEFTs in the future. This survey provides a valuable resource for both newcomers and experts seeking to understand and use the power of PEFT across FMs. All reviewed papers are listed at \url{https://github.com/THUDM/Awesome-Parameter-Efficient-Fine-Tuning-for-Foundation-Models}.
2501.13790
Local Steps Speed Up Local GD for Heterogeneous Distributed Logistic Regression
cs.LG
We analyze two variants of Local Gradient Descent applied to distributed logistic regression with heterogeneous, separable data and show convergence at the rate $O(1/KR)$ for $K$ local steps and sufficiently large $R$ communication rounds. In contrast, all existing convergence guarantees for Local GD applied to any problem are at least $\Omega(1/R)$, meaning they fail to show the benefit of local updates. The key to our improved guarantee is showing progress on the logistic regression objective when using a large stepsize $\eta \gg 1/K$, whereas prior analysis depends on $\eta \leq 1/K$.
2501.13794
Unveiling the Power of Noise Priors: Enhancing Diffusion Models for Mobile Traffic Prediction
cs.LG
Accurate prediction of mobile traffic, \textit{i.e.,} network traffic from cellular base stations, is crucial for optimizing network performance and supporting urban development. However, the non-stationary nature of mobile traffic, driven by human activity and environmental changes, leads to both regular patterns and abrupt variations. Diffusion models excel in capturing such complex temporal dynamics due to their ability to capture the inherent uncertainties. Most existing approaches prioritize designing novel denoising networks but often neglect the critical role of noise itself, potentially leading to sub-optimal performance. In this paper, we introduce a novel perspective by emphasizing the role of noise in the denoising process. Our analysis reveals that noise fundamentally shapes mobile traffic predictions, exhibiting distinct and consistent patterns. We propose NPDiff, a framework that decomposes noise into \textit{prior} and \textit{residual} components, with the \textit{prior} derived from data dynamics, enhancing the model's ability to capture both regular and abrupt variations. NPDiff can seamlessly integrate with various diffusion-based prediction models, delivering predictions that are effective, efficient, and robust. Extensive experiments demonstrate that it achieves superior performance with an improvement over 30\%, offering a new perspective on leveraging diffusion models in this domain.
2501.13795
Training-Free Zero-Shot Temporal Action Detection with Vision-Language Models
cs.CV
Existing zero-shot temporal action detection (ZSTAD) methods predominantly use fully supervised or unsupervised strategies to recognize unseen activities. However, these training-based methods are prone to domain shifts and require high computational costs, which hinder their practical applicability in real-world scenarios. In this paper, unlike previous works, we propose a training-Free Zero-shot temporal Action Detection (FreeZAD) method, leveraging existing vision-language (ViL) models to directly classify and localize unseen activities within untrimmed videos without any additional fine-tuning or adaptation. We mitigate the need for explicit temporal modeling and reliance on pseudo-label quality by designing the LOGarithmic decay weighted Outer-Inner-Contrastive Score (LogOIC) and frequency-based Actionness Calibration. Furthermore, we introduce a test-time adaptation (TTA) strategy using Prototype-Centric Sampling (PCS) to expand FreeZAD, enabling ViL models to adapt more effectively for ZSTAD. Extensive experiments on the THUMOS14 and ActivityNet-1.3 datasets demonstrate that our training-free method outperforms state-of-the-art unsupervised methods while requiring only 1/13 of the runtime. When equipped with TTA, the enhanced method further narrows the gap with fully supervised methods.
2501.13796
PromptMono: Cross Prompting Attention for Self-Supervised Monocular Depth Estimation in Challenging Environments
cs.CV
Considerable efforts have been made to improve monocular depth estimation under ideal conditions. However, in challenging environments, monocular depth estimation still faces difficulties. In this paper, we introduce visual prompt learning for predicting depth across different environments within a unified model, and present a self-supervised learning framework called PromptMono. It employs a set of learnable parameters as visual prompts to capture domain-specific knowledge. To integrate prompting information into image representations, a novel gated cross prompting attention (GCPA) module is proposed, which enhances the depth estimation in diverse conditions. We evaluate the proposed PromptMono on the Oxford Robotcar dataset and the nuScenes dataset. Experimental results demonstrate the superior performance of the proposed method.
2501.13804
Towards Real-World Validation of a Physics-Based Ship Motion Prediction Model
eess.SY cs.RO cs.SY
The maritime industry aims towards a sustainable future, which requires significant improvements in operational efficiency. Current approaches focus on minimising fuel consumption and emissions through greater autonomy. Efficient and safe autonomous navigation requires high-fidelity ship motion models applicable to real-world conditions. Although physics-based ship motion models can predict ships' motion with sub-second resolution, their validation in real-world conditions is rarely found in the literature. This study presents a physics-based 3D dynamics motion model that is tailored to a container-ship, and compares its predictions against real-world voyages. The model integrates vessel motion over time and accounts for its hydrodynamic behavior under different environmental conditions. The model's predictions are evaluated against real vessel data both visually and using multiple distance measures. Both methodologies demonstrate that the model's predictions align closely with the real-world trajectories of the container-ship.
2501.13805
EgoHand: Ego-centric Hand Pose Estimation and Gesture Recognition with Head-mounted Millimeter-wave Radar and IMUs
cs.CV
Recent advanced Virtual Reality (VR) headsets, such as the Apple Vision Pro, employ bottom-facing cameras to detect hand gestures and inputs, which offers users significant convenience in VR interactions. However, these bottom-facing cameras can sometimes be inconvenient and pose a risk of unintentionally exposing sensitive information, such as private body parts or personal surroundings. To mitigate these issues, we introduce EgoHand. This system provides an alternative solution by integrating millimeter-wave radar and IMUs for hand gesture recognition, thereby offering users an additional option for gesture interaction that enhances privacy protection. To accurately recognize hand gestures, we devise a two-stage skeleton-based gesture recognition scheme. In the first stage, a novel end-to-end Transformer architecture is employed to estimate the coordinates of hand joints. Subsequently, these estimated joint coordinates are utilized for gesture recognition. Extensive experiments involving 10 subjects show that EgoHand can detect hand gestures with 90.8% accuracy. Furthermore, EgoHand demonstrates robust performance across a variety of cross-domain tests, including different users, dominant hands, body postures, and scenes.
2501.13806
Generation of reusable learning objects from digital medical collections: An analysis based on the MASMDOA framework
cs.CL cs.HC
Learning Objects represent a widespread approach to structuring instructional materials in a large variety of educational contexts. The main aim of this work consists of analyzing from a qualitative point of view the process of generating reusable learning objects (RLOs) followed by Clavy, a tool that can be used to retrieve data from multiple medical knowledge sources and reconfigure such sources in diverse multimedia-based structures and organizations. From these organizations, Clavy is able to generate learning objects which can be adapted to various instructional healthcare scenarios with several types of user profiles and distinct learning requirements. Moreover, Clavy provides the capability of exporting these learning objects through educational standard specifications, which improves their reusability features. The analysis insights highlight the importance of having a tool able to transfer knowledge from the available digital medical collections to learning objects that can be easily accessed by medical students and healthcare practitioners through the most popular e-learning platforms.
2501.13810
Learning to Help in Multi-Class Settings
cs.LG cs.AI
Deploying complex machine learning models on resource-constrained devices is challenging due to limited computational power, memory, and model retrainability. To address these limitations, a hybrid system can be established by augmenting the local model with a server-side model, where samples are selectively deferred by a rejector and then sent to the server for processing. The hybrid system enables efficient use of computational resources while minimizing the overhead associated with server usage. The recently proposed Learning to Help (L2H) model trains a server model given a fixed local (client) model, differing from the Learning to Defer (L2D) framework, which trains the client for a fixed (expert) server. In both L2D and L2H, the training includes learning a rejector at the client to determine when to query the server. In this work, we extend the L2H model from binary to multi-class classification problems and demonstrate its applicability in a number of different scenarios of practical interest in which access to the server may be limited by cost, availability, or policy. We derive a stage-switching surrogate loss function that is differentiable, convex, and consistent with the Bayes rule corresponding to the 0-1 loss for the L2H model. Experiments show that our proposed methods offer an efficient and practical solution for multi-class classification in resource-constrained environments.
2501.13812
By-Example Synthesis of Vector Textures
cs.CV cs.GR
We propose a new method for synthesizing an arbitrarily sized novel vector texture given a single raster exemplar. Our method first segments the exemplar to extract the primary textons, and then clusters them based on visual similarity. We then compute a descriptor to capture each texton's neighborhood which contains the inter-category relationships that are used at synthesis time. Next, we use a simple procedure to both extract and place the secondary textons behind the primary polygons. Finally, our method constructs a gradient field for the background which is defined by a set of data points and colors. The color of the secondary polygons are also adjusted to better match the gradient field. To compare our work with other methods, we use a wide range of perceptual-based metrics.
2501.13814
On entropy-constrained Gaussian channel capacity via the moment problem
cs.IT math.IT math.PR
We study the capacity of the power-constrained additive Gaussian channel with an entropy constraint at the input. In particular, we characterize this capacity in the low signal-to-noise ratio regime, as a corollary of the following general result on a moment matching problem: we show that for any continuous random variable with finite moments, the largest number of initial moments that can be matched by a discrete random variable of sufficiently small but positive entropy is three.
2501.13816
Large Language Model driven Policy Exploration for Recommender Systems
cs.IR
Recent advancements in Recommender Systems (RS) have incorporated Reinforcement Learning (RL), framing the recommendation as a Markov Decision Process (MDP). However, offline RL policies trained on static user data are vulnerable to distribution shift when deployed in dynamic online environments. Additionally, excessive focus on exploiting short-term relevant items can hinder exploration, leading to suboptimal recommendations and negatively impacting long-term user gains. Online RL-based RS also face challenges in production deployment, due to the risks of exposing users to untrained or unstable policies. Large Language Models (LLMs) offer a promising solution to mimic user objectives and preferences for pre-training policies offline to enhance the initial recommendations in online settings. Effectively managing distribution shift and balancing exploration are crucial for improving RL-based RS, especially when leveraging LLM-based pre-training. To address these challenges, we propose an Interaction-Augmented Learned Policy (iALP) that utilizes user preferences distilled from an LLM. Our approach involves prompting the LLM with user states to extract item preferences, learning rewards based on feedback, and updating the RL policy using an actor-critic framework. Furthermore, to deploy iALP in an online scenario, we introduce an adaptive variant, A-iALP, that implements a simple fine-tuning strategy (A-iALP$_{ft}$), and an adaptive approach (A-iALP$_{ap}$) designed to mitigate issues with compromised policies and limited exploration. Experiments across three simulated environments demonstrate that A-iALP introduces substantial performance improvements
2501.13817
Temporal Logic Guided Safe Navigation for Autonomous Vehicles
cs.RO cs.FL cs.SY eess.SY
Safety verification for autonomous vehicles (AVs) and ground robots is crucial for ensuring reliable operation given their uncertain environments. Formal language tools provide a robust and sound method to verify safety rules for such complex cyber-physical systems. In this paper, we propose a hybrid approach that combines the strengths of formal verification languages like Linear Temporal Logic (LTL) and Signal Temporal Logic (STL) to generate safe trajectories and optimal control inputs for autonomous vehicle navigation. We implement a symbolic path planning approach using LTL to generate a formally safe reference trajectory. A mixed integer linear programming (MILP) solver is then used on this reference trajectory to solve for the control inputs while satisfying the state, control and safety constraints described by STL. We test our proposed solution on two environments and compare the results with popular path planning algorithms. In contrast to conventional path planning algorithms, our formally safe solution excels in handling complex specification scenarios while ensuring both safety and comparable computation times.
2501.13818
Ensuring Medical AI Safety: Explainable AI-Driven Detection and Mitigation of Spurious Model Behavior and Associated Data
cs.AI cs.CV cs.LG
Deep neural networks are increasingly employed in high-stakes medical applications, despite their tendency for shortcut learning in the presence of spurious correlations, which can have potentially fatal consequences in practice. Detecting and mitigating shortcut behavior is a challenging task that often requires significant labeling efforts from domain experts. To alleviate this problem, we introduce a semi-automated framework for the identification of spurious behavior from both data and model perspective by leveraging insights from eXplainable Artificial Intelligence (XAI). This allows the retrieval of spurious data points and the detection of model circuits that encode the associated prediction rules. Moreover, we demonstrate how these shortcut encodings can be used for XAI-based sample- and pixel-level data annotation, providing valuable information for bias mitigation methods to unlearn the undesired shortcut behavior. We show the applicability of our framework using four medical datasets across two modalities, featuring controlled and real-world spurious correlations caused by data artifacts. We successfully identify and mitigate these biases in VGG16, ResNet50, and contemporary Vision Transformer models, ultimately increasing their robustness and applicability for real-world medical tasks.
2501.13820
Consistent spectral clustering in sparse tensor block models
math.ST cs.LG math.PR stat.TH
High-order clustering aims to classify objects in multiway datasets that are prevalent in various fields such as bioinformatics, social network analysis, and recommendation systems. These tasks often involve data that is sparse and high-dimensional, presenting significant statistical and computational challenges. This paper introduces a tensor block model specifically designed for sparse integer-valued data tensors. We propose a simple spectral clustering algorithm augmented with a trimming step to mitigate noise fluctuations, and identify a density threshold that ensures the algorithm's consistency. Our approach models sparsity using a sub-Poisson noise concentration framework, accommodating heavier than sub-Gaussian tails. Remarkably, this natural class of tensor block models is closed under aggregation across arbitrary modes. Consequently, we obtain a comprehensive framework for evaluating the tradeoff between signal loss and noise reduction during data aggregation. The analysis is based on a novel concentration bound for sparse random Gram matrices. The theoretical findings are illustrated through simulation experiments.
2501.13824
Hallucinations Can Improve Large Language Models in Drug Discovery
cs.CL cs.AI
Concerns about hallucinations in Large Language Models (LLMs) have been raised by researchers, yet their potential in areas where creativity is vital, such as drug discovery, merits exploration. In this paper, we come up with the hypothesis that hallucinations can improve LLMs in drug discovery. To verify this hypothesis, we use LLMs to describe the SMILES string of molecules in natural language and then incorporate these descriptions as part of the prompt to address specific tasks in drug discovery. Evaluated on seven LLMs and five classification tasks, our findings confirm the hypothesis: LLMs can achieve better performance with text containing hallucinations. Notably, Llama-3.1-8B achieves an 18.35% gain in ROC-AUC compared to the baseline without hallucination. Furthermore, hallucinations generated by GPT-4o provide the most consistent improvements across models. Additionally, we conduct empirical analyses and a case study to investigate key factors affecting performance and the underlying reasons. Our research sheds light on the potential use of hallucinations for LLMs and offers new perspectives for future research leveraging LLMs in drug discovery.
2501.13825
Sample-Based Piecewise Linear Power Flow Approximations Using Second-Order Sensitivities
math.OC cs.SY eess.SY
The inherent nonlinearity of the power flow equations poses significant challenges in accurately modeling power systems, particularly when employing linearized approximations. Although power flow linearizations provide computational efficiency, they can fail to fully capture nonlinear behavior across diverse operating conditions. To improve approximation accuracy, we propose conservative piecewise linear approximations (CPLA) of the power flow equations, which are designed to consistently over- or under-estimate the quantity of interest, ensuring conservative behavior in optimization. The flexibility provided by piecewise linear functions can yield improved accuracy relative to standard linear approximations. However, applying CPLA across all dimensions of the power flow equations could introduce significant computational complexity, especially for large-scale optimization problems. In this paper, we propose a strategy that selectively targets dimensions exhibiting significant nonlinearities. Using a second-order sensitivity analysis, we identify the directions where the power flow equations exhibit the most significant curvature and tailor the CPLAs to improve accuracy in these specific directions. This approach reduces the computational burden while maintaining high accuracy, making it particularly well-suited for mixed-integer programming problems involving the power flow equations.
2501.13826
Video-MMMU: Evaluating Knowledge Acquisition from Multi-Discipline Professional Videos
cs.CV cs.CL
Humans acquire knowledge through three cognitive stages: perceiving information, comprehending knowledge, and adapting knowledge to solve novel problems. Videos serve as an effective medium for this learning process, facilitating a progression through these cognitive stages. However, existing video benchmarks fail to systematically evaluate the knowledge acquisition capabilities in Large Multimodal Models (LMMs). To address this gap, we introduce Video-MMMU, a multi-modal, multi-disciplinary benchmark designed to assess LMMs' ability to acquire and utilize knowledge from videos. Video-MMMU features a curated collection of 300 expert-level videos and 900 human-annotated questions across six disciplines, evaluating knowledge acquisition through stage-aligned question-answer pairs: Perception, Comprehension, and Adaptation. A proposed knowledge gain metric, {\Delta}knowledge, quantifies improvement in performance after video viewing. Evaluation of LMMs reveals a steep decline in performance as cognitive demands increase and highlights a significant gap between human and model knowledge acquisition, underscoring the need for methods to enhance LMMs' capability to learn and adapt from videos.
2501.13828
PhotoGAN: Generative Adversarial Neural Network Acceleration with Silicon Photonics
cs.AR cs.LG
Generative Adversarial Networks (GANs) are at the forefront of AI innovation, driving advancements in areas such as image synthesis, medical imaging, and data augmentation. However, the unique computational operations within GANs, such as transposed convolutions and instance normalization, introduce significant inefficiencies when executed on traditional electronic accelerators, resulting in high energy consumption and suboptimal performance. To address these challenges, we introduce PhotoGAN, the first silicon-photonic accelerator designed to handle the specialized operations of GAN models. By leveraging the inherent high throughput and energy efficiency of silicon photonics, PhotoGAN offers an innovative, reconfigurable architecture capable of accelerating transposed convolutions and other GAN-specific layers. The accelerator also incorporates a sparse computation optimization technique to reduce redundant operations, improving computational efficiency. Our experimental results demonstrate that PhotoGAN achieves at least 4.4x higher GOPS and 2.18x lower energy-per-bit (EPB) compared to state-of-the-art accelerators, including GPUs and TPUs. These findings showcase PhotoGAN as a promising solution for the next generation of GAN acceleration, providing substantial gains in both performance and energy efficiency.
2501.13829
MV-GMN: State Space Model for Multi-View Action Recognition
cs.CV
Recent advancements in multi-view action recognition have largely relied on Transformer-based models. While effective and adaptable, these models often require substantial computational resources, especially in scenarios with multiple views and multiple temporal sequences. Addressing this limitation, this paper introduces the MV-GMN model, a state-space model specifically designed to efficiently aggregate multi-modal data (RGB and skeleton), multi-view perspectives, and multi-temporal information for action recognition with reduced computational complexity. The MV-GMN model employs an innovative Multi-View Graph Mamba network comprising a series of MV-GMN blocks. Each block includes a proposed Bidirectional State Space Block and a GCN module. The Bidirectional State Space Block introduces four scanning strategies, including view-prioritized and time-prioritized approaches. The GCN module leverages rule-based and KNN-based methods to construct the graph network, effectively integrating features from different viewpoints and temporal instances. Demonstrating its efficacy, MV-GMN outperforms the state-of-the-arts on several datasets, achieving notable accuracies of 97.3\% and 96.7\% on the NTU RGB+D 120 dataset in cross-subject and cross-view scenarios, respectively. MV-GMN also surpasses Transformer-based baselines while requiring only linear inference complexity, underscoring the model's ability to reduce computational load and enhance the scalability and applicability of multi-view action recognition technologies.
2501.13830
A space-decoupling framework for optimization on bounded-rank matrices with orthogonally invariant constraints
math.OC cs.AI cs.LG
Imposing additional constraints on low-rank optimization has garnered growing interest. However, the geometry of coupled constraints hampers the well-developed low-rank structure and makes the problem intricate. To this end, we propose a space-decoupling framework for optimization on bounded-rank matrices with orthogonally invariant constraints. The ``space-decoupling" is reflected in several ways. We show that the tangent cone of coupled constraints is the intersection of tangent cones of each constraint. Moreover, we decouple the intertwined bounded-rank and orthogonally invariant constraints into two spaces, leading to optimization on a smooth manifold. Implementing Riemannian algorithms on this manifold is painless as long as the geometry of additional constraints is known. In addition, we unveil the equivalence between the reformulated problem and the original problem. Numerical experiments on real-world applications -- spherical data fitting, graph similarity measuring, low-rank SDP, model reduction of Markov processes, reinforcement learning, and deep learning -- validate the superiority of the proposed framework.
2501.13831
Predicting Compact Phrasal Rewrites with Large Language Models for ASR Post Editing
cs.CL cs.AI cs.LG
Large Language Models (LLMs) excel at rewriting tasks such as text style transfer and grammatical error correction. While there is considerable overlap between the inputs and outputs in these tasks, the decoding cost still increases with output length, regardless of the amount of overlap. By leveraging the overlap between the input and the output, Kaneko and Okazaki (2023) proposed model-agnostic edit span representations to compress the rewrites to save computation. They reported an output length reduction rate of nearly 80% with minimal accuracy impact in four rewriting tasks. In this paper, we propose alternative edit phrase representations inspired by phrase-based statistical machine translation. We systematically compare our phrasal representations with their span representations. We apply the LLM rewriting model to the task of Automatic Speech Recognition (ASR) post editing and show that our target-phrase-only edit representation has the best efficiency-accuracy trade-off. On the LibriSpeech test set, our method closes 50-60% of the WER gap between the edit span model and the full rewrite model while losing only 10-20% of the length reduction rate of the edit span model.
2501.13833
On the Reasoning Capacity of AI Models and How to Quantify It
cs.AI cs.CL cs.IT math.IT
Recent advances in Large Language Models (LLMs) have intensified the debate surrounding the fundamental nature of their reasoning capabilities. While achieving high performance on benchmarks such as GPQA and MMLU, these models exhibit limitations in more complex reasoning tasks, highlighting the need for more rigorous evaluation methodologies. We propose a novel phenomenological approach that goes beyond traditional accuracy metrics to probe the underlying mechanisms of model behavior, establishing a framework that could broadly impact how we analyze and understand AI systems. Using positional bias in multiple-choice reasoning tasks as a case study, we demonstrate how systematic perturbations can reveal fundamental aspects of model decision-making. To analyze these behaviors, we develop two complementary phenomenological models: a Probabilistic Mixture Model (PMM) that decomposes model responses into reasoning, memorization, and guessing components and an Information-Theoretic Consistency (ITC) analysis that quantifies the relationship between model confidence and strategy selection. Through controlled experiments on reasoning benchmarks, we show that true reasoning remains challenging for current models, with apparent success often relying on sophisticated combinations of memorization and pattern matching rather than genuine logical deduction. More fundamentally, we demonstrate that accuracy alone often overstates a model's reasoning abilities, as model behavior can be characterized through underlying mechanisms in the phase space of cognitive strategies, revealing how models dynamically balance different approaches when responding to queries. This framework enables quantitative criteria for real-world deployments, allowing applications to specify reliability thresholds based on strategy distributions rather than aggregate performance metrics.
2501.13836
Think Outside the Data: Colonial Biases and Systemic Issues in Automated Moderation Pipelines for Low-Resource Languages
cs.CL cs.HC
Most social media users come from non-English speaking countries in the Global South. Despite the widespread prevalence of harmful content in these regions, current moderation systems repeatedly struggle in low-resource languages spoken there. In this work, we examine the challenges AI researchers and practitioners face when building moderation tools for low-resource languages. We conducted semi-structured interviews with 22 AI researchers and practitioners specializing in automatic detection of harmful content in four diverse low-resource languages from the Global South. These are: Tamil from South Asia, Swahili from East Africa, Maghrebi Arabic from North Africa, and Quechua from South America. Our findings reveal that social media companies' restrictions on researchers' access to data exacerbate the historical marginalization of these languages, which have long lacked datasets for studying online harms. Moreover, common preprocessing techniques and language models, predominantly designed for data-rich English, fail to account for the linguistic complexity of low-resource languages. This leads to critical errors when moderating content in Tamil, Swahili, Arabic, and Quechua, which are morphologically richer than English. Based on our findings, we establish that the precarities in current moderation pipelines are rooted in deep systemic inequities and continue to reinforce historical power imbalances. We conclude by discussing multi-stakeholder approaches to improve moderation for low-resource languages.
2501.13848
Where Do You Go? Pedestrian Trajectory Prediction using Scene Features
cs.CV cs.AI cs.LG
Accurate prediction of pedestrian trajectories is crucial for enhancing the safety of autonomous vehicles and reducing traffic fatalities involving pedestrians. While numerous studies have focused on modeling interactions among pedestrians to forecast their movements, the influence of environmental factors and scene-object placements has been comparatively underexplored. In this paper, we present a novel trajectory prediction model that integrates both pedestrian interactions and environmental context to improve prediction accuracy. Our approach captures spatial and temporal interactions among pedestrians within a sparse graph framework. To account for pedestrian-scene interactions, we employ advanced image enhancement and semantic segmentation techniques to extract detailed scene features. These scene and interaction features are then fused through a cross-attention mechanism, enabling the model to prioritize relevant environmental factors that influence pedestrian movements. Finally, a temporal convolutional network processes the fused features to predict future pedestrian trajectories. Experimental results demonstrate that our method significantly outperforms existing state-of-the-art approaches, achieving ADE and FDE values of 0.252 and 0.372 meters, respectively, underscoring the importance of incorporating both social interactions and environmental context in pedestrian trajectory prediction.
2501.13851
Large Vision-Language Models for Knowledge-Grounded Data Annotation of Memes
cs.LG
Memes have emerged as a powerful form of communication, integrating visual and textual elements to convey humor, satire, and cultural messages. Existing research has focused primarily on aspects such as emotion classification, meme generation, propagation, interpretation, figurative language, and sociolinguistics, but has often overlooked deeper meme comprehension and meme-text retrieval. To address these gaps, this study introduces ClassicMemes-50-templates (CM50), a large-scale dataset consisting of over 33,000 memes, centered around 50 popular meme templates. We also present an automated knowledge-grounded annotation pipeline leveraging large vision-language models to produce high-quality image captions, meme captions, and literary device labels overcoming the labor intensive demands of manual annotation. Additionally, we propose a meme-text retrieval CLIP model (mtrCLIP) that utilizes cross-modal embedding to enhance meme analysis, significantly improving retrieval performance. Our contributions include:(1) a novel dataset for large-scale meme study, (2) a scalable meme annotation framework, and (3) a fine-tuned CLIP for meme-text retrieval, all aimed at advancing the understanding and analysis of memes at scale.
2501.13855
First Lessons Learned of an Artificial Intelligence Robotic System for Autonomous Coarse Waste Recycling Using Multispectral Imaging-Based Methods
cs.CV cs.LG cs.RO
Current disposal facilities for coarse-grained waste perform manual sorting of materials with heavy machinery. Large quantities of recyclable materials are lost to coarse waste, so more effective sorting processes must be developed to recover them. Two key aspects to automate the sorting process are object detection with material classification in mixed piles of waste, and autonomous control of hydraulic machinery. Because most objects in those accumulations of waste are damaged or destroyed, object detection alone is not feasible in the majority of cases. To address these challenges, we propose a classification of materials with multispectral images of ultraviolet (UV), visual (VIS), near infrared (NIR), and short-wave infrared (SWIR) spectrums. Solution for autonomous control of hydraulic heavy machines for sorting of bulky waste is being investigated using cost-effective cameras and artificial intelligence-based controllers.
2501.13858
The Lock Generative Adversarial Network for Medical Waveform Anomaly Detection
cs.CE
Waveform signal analysis is a complex and important task in medical care. For example, mechanical ventilators are critical life-support machines, but they can cause serious injury to patients if they are out of synchronization with the patients' own breathing reflex. This asynchrony is revealed by the waveforms showing flow and pressure histories. Likewise, electrocardiograms record the electrical activity of a patients' heart as a set of waveforms, and anomalous waveforms can reveal important disease states. In both cases, subtle variations in a complex waveform are important information for patient care; signals which may be missed or mis-interpreted by human caregivers. We report on the design of a novel Lock Generative Adversarial Network architecture for anomaly detection in raw or summarized medical waveform data. The proposed architecture uses alternating optimization of the generator and discriminator networks to solve the convergence dilemma. Furthermore, the fidelity of the generator networks' outputs to the actual distribution of anomalous data is improved via synthetic minority oversampling. We evaluate this new architecture on one ventilator asynchrony dataset, and two electrocardiogram datasets, finding that the performance was either equal or superior to the state-of-the art on all three.
2501.13859
Dual-Modal Prototype Joint Learning for Compositional Zero-Shot Learning
cs.CV
Compositional Zero-Shot Learning (CZSL) aims to recognize novel compositions of attributes and objects by leveraging knowledge learned from seen compositions. Recent approaches have explored the use of Vision-Language Models (VLMs) to align textual and visual modalities. These methods typically employ prompt engineering, parameter-tuning, and modality fusion to generate rich textual prototypes that serve as class prototypes for CZSL. However, the modality gap results in textual prototypes being unable to fully capture the optimal representations of all class prototypes, particularly those with fine-grained features, which can be directly obtained from the visual modality. In this paper, we propose a novel Dual-Modal Prototype Joint Learning framework for the CZSL task. Our approach, based on VLMs, introduces prototypes in both the textual and visual modalities. The textual prototype is optimized to capture broad conceptual information, aiding the model's generalization across unseen compositions. Meanwhile, the visual prototype is used to mitigate the classification errors caused by the modality gap and capture fine-grained details to distinguish images with similar appearances. To effectively optimize these prototypes, we design specialized decomposition modules and a joint learning strategy that enrich the features from both modalities. These prototypes not only capture key category information during training but also serve as crucial reference targets during inference. Experimental results demonstrate that our approach achieves state-of-the-art performance in the closed-world setting and competitive performance in the open-world setting across three publicly available CZSL benchmarks. These findings validate the effectiveness of our method in advancing compositional generalization.
2501.13864
Autoencoders for Anomaly Detection are Unreliable
cs.LG cs.AI
Autoencoders are frequently used for anomaly detection, both in the unsupervised and semi-supervised settings. They rely on the assumption that when trained using the reconstruction loss, they will be able to reconstruct normal data more accurately than anomalous data. Some recent works have posited that this assumption may not always hold, but little has been done to study the validity of the assumption in theory. In this work we show that this assumption indeed does not hold, and illustrate that anomalies, lying far away from normal data, can be perfectly reconstructed in practice. We revisit the theory of failure of linear autoencoders for anomaly detection by showing how they can perfectly reconstruct out of bounds, or extrapolate undesirably, and note how this can be dangerous in safety critical applications. We connect this to non-linear autoencoders through experiments on both tabular data and real-world image data, the two primary application areas of autoencoders for anomaly detection.
2501.13865
Threshold Selection for Iterative Decoding of $(v,w)$-regular Binary Codes
cs.CR cs.IT math.IT
Iterative bit flipping decoders are an efficient and effective decoder choice for decoding codes which admit a sparse parity-check matrix. Among these, sparse $(v,w)$-regular codes, which include LDPC and MDPC codes are of particular interest both for efficient data correction and the design of cryptographic primitives. In attaining the decoding the choice of the bit flipping thresholds, which can be determined either statically, or during the decoder execution by using information coming from the initial syndrome value and its updates. In this work, we analyze a two-iterations parallel hard decision bit flipping decoders and propose concrete criteria for threshold determination, backed by a closed form model. In doing so, we introduce a new tightly fitting model for the distribution of the Hamming weight of the syndrome after the first decoder iteration and substantial improvements on the DFR estimation with respect to existing approaches.
2501.13868
Lost in Siting: The Hidden Carbon Cost of Inequitable Residential Solar Installations
cs.CE
The declining cost of solar photovoltaics (PV) combined with strong federal and state-level incentives have resulted in a high number of residential solar PV installations in the US. However, these installations are concentrated in particular regions, such as California, and demographics, such as high-income Asian neighborhoods. This inequitable distribution creates an illusion that further increasing residential solar installations will become increasingly challenging. Furthermore, while the inequity in solar installations has received attention, no prior comprehensive work has been done on understanding whether our current trajectory of residential solar adoption is energy- and carbon-efficient. In this paper, we reveal the hidden energy and carbon cost of the inequitable distribution of existing installations. Using US-based data on carbon offset potential, the amount of avoided carbon emissions from using rooftop PV instead of electric grid energy, and the number of existing solar installations, we surprisingly observe that locations and demographics with a higher carbon offset potential have fewer existing installations. For instance, neighborhoods with relatively higher black population have 7.4% higher carbon offset potential than average but 36.7% fewer installations; lower-income neighborhoods have 14.7% higher potential and 47% fewer installations. We propose several equity- and carbon-aware solar siting strategies. In evaluating these strategies, we develop Sunsight, a toolkit that combines simulation/visualization tools and our relevant datasets, which we are releasing publicly. Our projections show that a multi-objective siting strategy can address two problems at once; namely, it can improve societal outcomes in terms of distributional equity and simultaneously improve the carbon-efficiency (i.e., climate impact) of current installation trends by up to 39.8%.
2501.13876
FAST-LIVO2 on Resource-Constrained Platforms: LiDAR-Inertial-Visual Odometry with Efficient Memory and Computation
cs.RO
This paper presents a lightweight LiDAR-inertial-visual odometry system optimized for resource-constrained platforms. It integrates a degeneration-aware adaptive visual frame selector into error-state iterated Kalman filter (ESIKF) with sequential updates, improving computation efficiency significantly while maintaining a similar level of robustness. Additionally, a memory-efficient mapping structure combining a locally unified visual-LiDAR map and a long-term visual map achieves a good trade-off between performance and memory usage. Extensive experiments on x86 and ARM platforms demonstrate the system's robustness and efficiency. On the Hilti dataset, our system achieves a 33% reduction in per-frame runtime and 47% lower memory usage compared to FAST-LIVO2, with only a 3 cm increase in RMSE. Despite this slight accuracy trade-off, our system remains competitive, outperforming state-of-the-art (SOTA) LIO methods such as FAST-LIO2 and most existing LIVO systems. These results validate the system's capability for scalable deployment on resource-constrained edge computing platforms.
2501.13878
Eye Gaze as a Signal for Conveying User Attention in Contextual AI Systems
cs.HC cs.CV
Advanced multimodal AI agents can now collaborate with users to solve challenges in the world. We explore eye tracking's role in such interaction to convey a user's attention relative to the physical environment. We hypothesize that this knowledge improves contextual understanding for AI agents. By observing hours of human-object interactions, we first measure the relationship between an eye tracker's signal quality and its ability to reliably place gaze on nearby physical objects. We then conduct experiments which relay the user's scanpath history as additional context querying multimodal agents. Our results show that eye tracking provides high value as a user attention signal and can convey information about the user's current task and interests to the agent.
2501.13880
A RAG-Based Institutional Assistant
cs.CL
Although large language models (LLMs) demonstrate strong text generation capabilities, they struggle in scenarios requiring access to structured knowledge bases or specific documents, limiting their effectiveness in knowledge-intensive tasks. To address this limitation, retrieval-augmented generation (RAG) models have been developed, enabling generative models to incorporate relevant document fragments into their inputs. In this paper, we design and evaluate a RAG-based virtual assistant specifically tailored for the University of S\~ao Paulo. Our system architecture comprises two key modules: a retriever and a generative model. We experiment with different types of models for both components, adjusting hyperparameters such as chunk size and the number of retrieved documents. Our optimal retriever model achieves a Top-5 accuracy of 30%, while our most effective generative model scores 22.04\% against ground truth answers. Notably, when the correct document chunks are supplied to the LLMs, accuracy significantly improves to 54.02%, an increase of over 30 percentage points. Conversely, without contextual input, performance declines to 13.68%. These findings highlight the critical role of database access in enhancing LLM performance. They also reveal the limitations of current semantic search methods in accurately identifying relevant documents and underscore the ongoing challenges LLMs face in generating precise responses.
2501.13883
Utilizing Evolution Strategies to Train Transformers in Reinforcement Learning
cs.LG cs.NE
We explore a capability of evolution strategies to train an agent with its policy based on a transformer architecture in a reinforcement learning setting. We performed experiments using OpenAI's highly parallelizable evolution strategy to train Decision Transformer in Humanoid locomotion environment and in the environment of Atari games, testing the ability of this black-box optimization technique to train even such relatively large and complicated models (compared to those previously tested in the literature). We also proposed a method to aid the training by first pretraining the model before using the OpenAI-ES to train it further, and tested its effectiveness. The examined evolution strategy proved to be, in general, capable of achieving strong results and managed to obtain high-performing agents. Therefore, the pretraining was shown to be unnecessary; yet still, it helped us observe and formulate several further insights.
2501.13884
Exploring Finetuned Audio-LLM on Heart Murmur Features
eess.AS cs.AI cs.SD
Large language models (LLMs) for audio have excelled in recognizing and analyzing human speech, music, and environmental sounds. However, their potential for understanding other types of sounds, particularly biomedical sounds, remains largely underexplored despite significant scientific interest. In this study, we focus on diagnosing cardiovascular diseases using phonocardiograms, i.e., heart sounds. Most existing deep neural network (DNN) paradigms are restricted to heart murmur classification (healthy vs unhealthy) and do not predict other acoustic features of the murmur such as timing, grading, harshness, pitch, and quality, which are important in helping physicians diagnose the underlying heart conditions. We propose to finetune an audio LLM, Qwen2-Audio, on the PhysioNet CirCor DigiScope phonocardiogram (PCG) dataset and evaluate its performance in classifying 11 expert-labeled murmur features. Additionally, we aim to achieve more noise-robust and generalizable system by exploring a preprocessing segmentation algorithm using an audio representation model, SSAMBA. Our results indicate that the LLM-based model outperforms state-of-the-art methods in 8 of the 11 features and performs comparably in the remaining 3. Moreover, the LLM successfully classifies long-tail murmur features with limited training data, a task that all previous methods have failed to classify. These findings underscore the potential of audio LLMs as assistants to human cardiologists in enhancing heart disease diagnosis.
2501.13885
Quantum model reduction for continuous-time quantum filters
quant-ph cs.SY eess.SY math-ph math.MP
The use of quantum stochastic models is widespread in dynamical reduction, simulation of open systems, feedback control and adaptive estimation. In many applications only part of the information contained in the filter's state is actually needed to reconstruct the target observable quantities; thus, filters of smaller dimensions could be in principle implemented to perform the same task.In this work, we propose a systematic method to find, when possible, reduced-order quantum filters that are capable of exactly reproducing the evolution of expectation values of interest. In contrast with existing reduction techniques, the reduced model we obtain is exact and in the form of a Belavkin filtering equation, ensuring physical interpretability.This is attained by leveraging tools from the theory of both minimal realization and non-commutative conditional expectations. The proposed procedure is tested on prototypical examples, laying the groundwork for applications in quantum trajectory simulation and quantum feedback control.
2501.13887
What Does an Audio Deepfake Detector Focus on? A Study in the Time Domain
cs.LG cs.SD eess.AS
Adding explanations to audio deepfake detection (ADD) models will boost their real-world application by providing insight on the decision making process. In this paper, we propose a relevancy-based explainable AI (XAI) method to analyze the predictions of transformer-based ADD models. We compare against standard Grad-CAM and SHAP-based methods, using quantitative faithfulness metrics as well as a partial spoof test, to comprehensively analyze the relative importance of different temporal regions in an audio. We consider large datasets, unlike previous works where only limited utterances are studied, and find that the XAI methods differ in their explanations. The proposed relevancy-based XAI method performs the best overall on a variety of metrics. Further investigation on the relative importance of speech/non-speech, phonetic content, and voice onsets/offsets suggest that the XAI results obtained from analyzing limited utterances don't necessarily hold when evaluated on large datasets.
2501.13888
Multimodal Sensor Dataset for Monitoring Older Adults Post Lower-Limb Fractures in Community Settings
cs.LG cs.CV
Lower-Limb Fractures (LLF) are a major health concern for older adults, often leading to reduced mobility and prolonged recovery, potentially impairing daily activities and independence. During recovery, older adults frequently face social isolation and functional decline, complicating rehabilitation and adversely affecting physical and mental health. Multi-modal sensor platforms that continuously collect data and analyze it using machine-learning algorithms can remotely monitor this population and infer health outcomes. They can also alert clinicians to individuals at risk of isolation and decline. This paper presents a new publicly available multi-modal sensor dataset, MAISON-LLF, collected from older adults recovering from LLF in community settings. The dataset includes data from smartphone and smartwatch sensors, motion detectors, sleep-tracking mattresses, and clinical questionnaires on isolation and decline. The dataset was collected from ten older adults living alone at home for eight weeks each, totaling 560 days of 24-hour sensor data. For technical validation, supervised machine-learning and deep-learning models were developed using the sensor and clinical questionnaire data, providing a foundational comparison for the research community.
2501.13889
Generating Realistic Forehead-Creases for User Verification via Conditioned Piecewise Polynomial Curves
cs.CV
We propose a trait-specific image generation method that models forehead creases geometrically using B-spline and B\'ezier curves. This approach ensures the realistic generation of both principal creases and non-prominent crease patterns, effectively constructing detailed and authentic forehead-crease images. These geometrically rendered images serve as visual prompts for a diffusion-based Edge-to-Image translation model, which generates corresponding mated samples. The resulting novel synthetic identities are then used to train a forehead-crease verification network. To enhance intra-subject diversity in the generated samples, we employ two strategies: (a) perturbing the control points of B-splines under defined constraints to maintain label consistency, and (b) applying image-level augmentations to the geometric visual prompts, such as dropout and elastic transformations, specifically tailored to crease patterns. By integrating the proposed synthetic dataset with real-world data, our method significantly improves the performance of forehead-crease verification systems under a cross-database verification protocol.
2501.13890
Federated Granger Causality Learning for Interdependent Clients with State Space Representation
cs.LG stat.ML
Advanced sensors and IoT devices have improved the monitoring and control of complex industrial enterprises. They have also created an interdependent fabric of geographically distributed process operations (clients) across these enterprises. Granger causality is an effective approach to detect and quantify interdependencies by examining how one client's state affects others over time. Understanding these interdependencies captures how localized events, such as faults and disruptions, can propagate throughout the system, possibly causing widespread operational impacts. However, the large volume and complexity of industrial data pose challenges in modeling these interdependencies. This paper develops a federated approach to learning Granger causality. We utilize a linear state space system framework that leverages low-dimensional state estimates to analyze interdependencies. This addresses bandwidth limitations and the computational burden commonly associated with centralized data processing. We propose augmenting the client models with the Granger causality information learned by the server through a Machine Learning (ML) function. We examine the co-dependence between the augmented client and server models and reformulate the framework as a standalone ML algorithm providing conditions for its sublinear and linear convergence rates. We also study the convergence of the framework to a centralized oracle model. Moreover, we include a differential privacy analysis to ensure data security while preserving causal insights. Using synthetic data, we conduct comprehensive experiments to demonstrate the robustness of our approach to perturbations in causality, the scalability to the size of communication, number of clients, and the dimensions of raw data. We also evaluate the performance on two real-world industrial control system datasets by reporting the volume of data saved by decentralization.
2501.13893
Pix2Cap-COCO: Advancing Visual Comprehension via Pixel-Level Captioning
cs.CV cs.AI cs.LG
We present Pix2Cap-COCO, the first panoptic pixel-level caption dataset designed to advance fine-grained visual understanding. To achieve this, we carefully design an automated annotation pipeline that prompts GPT-4V to generate pixel-aligned, instance-specific captions for individual objects within images, enabling models to learn more granular relationships between objects and their contexts. This approach results in 167,254 detailed captions, with an average of 22.94 words per caption. Building on Pix2Cap-COCO, we introduce a novel task, panoptic segmentation-captioning, which challenges models to recognize instances in an image and provide detailed descriptions for each simultaneously. To benchmark this task, we design a robust baseline based on X-Decoder. The experimental results demonstrate that Pix2Cap-COCO is a particularly challenging dataset, as it requires models to excel in both fine-grained visual understanding and detailed language generation. Furthermore, we leverage Pix2Cap-COCO for Supervised Fine-Tuning (SFT) on large multimodal models (LMMs) to enhance their performance. For example, training with Pix2Cap-COCO significantly improves the performance of GPT4RoI, yielding gains in CIDEr +1.4%, ROUGE +0.4%, and SPICE +0.5% on Visual Genome dataset, and strengthens its region understanding ability on the ViP-BENCH, with an overall improvement of +5.1%, including notable increases in recognition accuracy +11.2% and language generation quality +22.2%.
2501.13896
GUI-Bee: Align GUI Action Grounding to Novel Environments via Autonomous Exploration
cs.CL cs.AI cs.CV cs.LG
Graphical User Interface (GUI) action grounding is a critical step in GUI automation that maps language instructions to actionable elements on GUI screens. Most recent works of GUI action grounding leverage large GUI datasets to fine-tune MLLMs. However, the fine-tuning data always covers limited GUI environments, and we find the performance of the resulting model deteriorates in novel environments. We argue that the GUI grounding models should be further aligned to the novel environments to reveal their full potential, when the inference is known to involve novel environments, i.e., environments not used during the previous fine-tuning. To realize this, we first propose GUI-Bee, an MLLM-based autonomous agent, to collect high-quality, environment-specific data through exploration and then continuously fine-tune GUI grounding models with the collected data. Our agent leverages a novel Q-value-Incentive In-Context Reinforcement Learning (Q-ICRL) method to optimize exploration efficiency and data quality. Additionally, we introduce NovelScreenSpot, a benchmark for testing how well the data can help align GUI action grounding models to novel environments and demonstrate the effectiveness of data collected by GUI-Bee in the experiments. Furthermore, we conduct an ablation study to validate the Q-ICRL method in enhancing the efficiency of GUI-Bee. Project page: https://gui-bee.github.io
2501.13898
PointOBB-v3: Expanding Performance Boundaries of Single Point-Supervised Oriented Object Detection
cs.CV cs.AI
With the growing demand for oriented object detection (OOD), recent studies on point-supervised OOD have attracted significant interest. In this paper, we propose PointOBB-v3, a stronger single point-supervised OOD framework. Compared to existing methods, it generates pseudo rotated boxes without additional priors and incorporates support for the end-to-end paradigm. PointOBB-v3 functions by integrating three unique image views: the original view, a resized view, and a rotated/flipped (rot/flp) view. Based on the views, a scale augmentation module and an angle acquisition module are constructed. In the first module, a Scale-Sensitive Consistency (SSC) loss and a Scale-Sensitive Feature Fusion (SSFF) module are introduced to improve the model's ability to estimate object scale. To achieve precise angle predictions, the second module employs symmetry-based self-supervised learning. Additionally, we introduce an end-to-end version that eliminates the pseudo-label generation process by integrating a detector branch and introduces an Instance-Aware Weighting (IAW) strategy to focus on high-quality predictions. We conducted extensive experiments on the DIOR-R, DOTA-v1.0/v1.5/v2.0, FAIR1M, STAR, and RSAR datasets. Across all these datasets, our method achieves an average improvement in accuracy of 3.56% in comparison to previous state-of-the-art methods. The code will be available at https://github.com/ZpyWHU/PointOBB-v3.
2501.13904
Privacy-Preserving Personalized Federated Prompt Learning for Multimodal Large Language Models
cs.LG
Multimodal Large Language Models (LLMs) are pivotal in revolutionizing customer support and operations by integrating multiple modalities such as text, images, and audio. Federated Prompt Learning (FPL) is a recently proposed approach that combines pre-trained multimodal LLMs such as vision-language models with federated learning to create personalized, privacy-preserving AI systems. However, balancing the competing goals of personalization, generalization, and privacy remains a significant challenge. Over-personalization can lead to overfitting, reducing generalizability, while stringent privacy measures, such as differential privacy, can hinder both personalization and generalization. In this paper, we propose a Differentially Private Federated Prompt Learning (DP-FPL) approach to tackle this challenge by leveraging a low-rank factorization scheme to capture generalization while maintaining a residual term that preserves expressiveness for personalization. To ensure privacy, we introduce a novel method where we apply local differential privacy to the two low-rank components of the local prompt, and global differential privacy to the global prompt. Our approach mitigates the impact of privacy noise on the model performance while balancing the tradeoff between personalization and generalization. Extensive experiments demonstrate the effectiveness of our approach over other benchmarks.
2501.13905
On Learning Representations for Tabular Data Distillation
cs.LG
Dataset distillation generates a small set of information-rich instances from a large dataset, resulting in reduced storage requirements, privacy or copyright risks, and computational costs for downstream modeling, though much of the research has focused on the image data modality. We study tabular data distillation, which brings in novel challenges such as the inherent feature heterogeneity and the common use of non-differentiable learning models (such as decision tree ensembles and nearest-neighbor predictors). To mitigate these challenges, we present $\texttt{TDColER}$, a tabular data distillation framework via column embeddings-based representation learning. To evaluate this framework, we also present a tabular data distillation benchmark, ${{\sf \small TDBench}}$. Based on an elaborate evaluation on ${{\sf \small TDBench}}$, resulting in 226,890 distilled datasets and 548,880 models trained on them, we demonstrate that $\texttt{TDColER}$ is able to boost the distilled data quality of off-the-shelf distillation schemes by 0.5-143% across 7 different tabular learning models.
2501.13906
Universal optimality of $T$-avoiding spherical codes and designs
math.CO cs.IT math.IT math.MG
Given an open set (a union of open intervals), $T\subset [-1,1]$ we introduce the concepts of $T$-avoiding spherical codes and designs, that is, spherical codes that have no inner products in the set $T$. We show that certain codes found in the minimal vectors of the Leech lattices, as well as the minimal vectors of the Barnes--Wall lattice and codes derived from strongly regular graphs, are universally optimal in the restricted class of $T$-avoiding codes. We also extend a result of Delsarte--Goethals--Seidel about codes with three inner products $\alpha, \beta, \gamma$ (in our terminology $(\alpha,\beta)$-avoiding $\gamma$-codes). Parallel to the notion of tight spherical designs, we also derive that these codes are minimal (tight) $T$-avoiding spherical designs of fixed dimension and strength. In some cases, we also find that codes under consideration have maximal cardinality in their $T$-avoiding class for given dimension and minimum distance.
2501.13908
Graph Neural Controlled Differential Equations For Collaborative Filtering
cs.IR
Graph Convolution Networks (GCNs) are widely considered state-of-the-art for recommendation systems. Several studies in the field of recommendation systems have attempted to apply collaborative filtering (CF) into the Neural ODE framework. These studies follow the same idea as LightGCN, which removes the weight matrix or with a discrete weight matrix. However, we argue that weight control is critical for neural ODE-based methods. The importance of weight in creating tailored graph convolution for each node is crucial, and employing a fixed/discrete weight means it cannot adjust over time within the ODE function. This rigidity in the graph convolution reduces its adaptability, consequently hindering the performance of recommendations. In this study, to create an optimal control for Neural ODE-based recommendation, we introduce a new method called Graph Neural Controlled Differential Equations for Collaborative Filtering (CDE-CF). Our method improves the performance of the Graph ODE-based method by incorporating weight control in a continuous manner. To evaluate our approach, we conducted experiments on various datasets. The results show that our method surpasses competing baselines, including GCNs-based models and state-of-the-art Graph ODE-based methods.
2501.13912
Analysis of Indic Language Capabilities in LLMs
cs.CL
This report evaluates the performance of text-in text-out Large Language Models (LLMs) to understand and generate Indic languages. This evaluation is used to identify and prioritize Indic languages suited for inclusion in safety benchmarks. We conduct this study by reviewing existing evaluation studies and datasets; and a set of twenty-eight LLMs that support Indic languages. We analyze the LLMs on the basis of the training data, license for model and data, type of access and model developers. We also compare Indic language performance across evaluation datasets and find that significant performance disparities in performance across Indic languages. Hindi is the most widely represented language in models. While model performance roughly correlates with number of speakers for the top five languages, the assessment after that varies.
2501.13915
Binary Diffusion Probabilistic Model
cs.CV
We introduce the Binary Diffusion Probabilistic Model (BDPM), a novel generative model optimized for binary data representations. While denoising diffusion probabilistic models (DDPMs) have demonstrated notable success in tasks like image synthesis and restoration, traditional DDPMs rely on continuous data representations and mean squared error (MSE) loss for training, applying Gaussian noise models that may not be optimal for discrete or binary data structures. BDPM addresses this by decomposing images into bitplanes and employing XOR-based noise transformations, with a denoising model trained using binary cross-entropy loss. This approach enables precise noise control and computationally efficient inference, significantly lowering computational costs and improving model convergence. When evaluated on image restoration tasks such as image super-resolution, inpainting, and blind image restoration, BDPM outperforms state-of-the-art methods on the FFHQ, CelebA, and CelebA-HQ datasets. Notably, BDPM requires fewer inference steps than traditional DDPM models to reach optimal results, showcasing enhanced inference efficiency.
2501.13916
PBM-VFL: Vertical Federated Learning with Feature and Sample Privacy
cs.LG
We present Poisson Binomial Mechanism Vertical Federated Learning (PBM-VFL), a communication-efficient Vertical Federated Learning algorithm with Differential Privacy guarantees. PBM-VFL combines Secure Multi-Party Computation with the recently introduced Poisson Binomial Mechanism to protect parties' private datasets during model training. We define the novel concept of feature privacy and analyze end-to-end feature and sample privacy of our algorithm. We compare sample privacy loss in VFL with privacy loss in HFL. We also provide the first theoretical characterization of the relationship between privacy budget, convergence error, and communication cost in differentially-private VFL. Finally, we empirically show that our model performs well with high levels of privacy.
2501.13918
Improving Video Generation with Human Feedback
cs.CV cs.AI cs.GR cs.LG
Video generation has achieved significant advances through rectified flow techniques, but issues like unsmooth motion and misalignment between videos and prompts persist. In this work, we develop a systematic pipeline that harnesses human feedback to mitigate these problems and refine the video generation model. Specifically, we begin by constructing a large-scale human preference dataset focused on modern video generation models, incorporating pairwise annotations across multi-dimensions. We then introduce VideoReward, a multi-dimensional video reward model, and examine how annotations and various design choices impact its rewarding efficacy. From a unified reinforcement learning perspective aimed at maximizing reward with KL regularization, we introduce three alignment algorithms for flow-based models by extending those from diffusion models. These include two training-time strategies: direct preference optimization for flow (Flow-DPO) and reward weighted regression for flow (Flow-RWR), and an inference-time technique, Flow-NRG, which applies reward guidance directly to noisy videos. Experimental results indicate that VideoReward significantly outperforms existing reward models, and Flow-DPO demonstrates superior performance compared to both Flow-RWR and standard supervised fine-tuning methods. Additionally, Flow-NRG lets users assign custom weights to multiple objectives during inference, meeting personalized video quality needs. Project page: https://gongyeliu.github.io/videoalign.
2501.13919
Temporal Preference Optimization for Long-Form Video Understanding
cs.CV cs.AI cs.CL cs.LG cs.RO
Despite significant advancements in video large multimodal models (video-LMMs), achieving effective temporal grounding in long-form videos remains a challenge for existing models. To address this limitation, we propose Temporal Preference Optimization (TPO), a novel post-training framework designed to enhance the temporal grounding capabilities of video-LMMs through preference learning. TPO adopts a self-training approach that enables models to differentiate between well-grounded and less accurate temporal responses by leveraging curated preference datasets at two granularities: localized temporal grounding, which focuses on specific video segments, and comprehensive temporal grounding, which captures extended temporal dependencies across entire video sequences. By optimizing on these preference datasets, TPO significantly enhances temporal understanding while reducing reliance on manually annotated data. Extensive experiments on three long-form video understanding benchmarks--LongVideoBench, MLVU, and Video-MME--demonstrate the effectiveness of TPO across two state-of-the-art video-LMMs. Notably, LLaVA-Video-TPO establishes itself as the leading 7B model on the Video-MME benchmark, underscoring the potential of TPO as a scalable and efficient solution for advancing temporal reasoning in long-form video understanding. Project page: https://ruili33.github.io/tpo_website.
2501.13920
IMAGINE-E: Image Generation Intelligence Evaluation of State-of-the-art Text-to-Image Models
cs.CV cs.CL cs.LG
With the rapid development of diffusion models, text-to-image(T2I) models have made significant progress, showcasing impressive abilities in prompt following and image generation. Recently launched models such as FLUX.1 and Ideogram2.0, along with others like Dall-E3 and Stable Diffusion 3, have demonstrated exceptional performance across various complex tasks, raising questions about whether T2I models are moving towards general-purpose applicability. Beyond traditional image generation, these models exhibit capabilities across a range of fields, including controllable generation, image editing, video, audio, 3D, and motion generation, as well as computer vision tasks like semantic segmentation and depth estimation. However, current evaluation frameworks are insufficient to comprehensively assess these models' performance across expanding domains. To thoroughly evaluate these models, we developed the IMAGINE-E and tested six prominent models: FLUX.1, Ideogram2.0, Midjourney, Dall-E3, Stable Diffusion 3, and Jimeng. Our evaluation is divided into five key domains: structured output generation, realism, and physical consistency, specific domain generation, challenging scenario generation, and multi-style creation tasks. This comprehensive assessment highlights each model's strengths and limitations, particularly the outstanding performance of FLUX.1 and Ideogram2.0 in structured and specific domain tasks, underscoring the expanding applications and potential of T2I models as foundational AI tools. This study provides valuable insights into the current state and future trajectory of T2I models as they evolve towards general-purpose usability. Evaluation scripts will be released at https://github.com/jylei16/Imagine-e.
2501.13921
The Breeze 2 Herd of Models: Traditional Chinese LLMs Based on Llama with Vision-Aware and Function-Calling Capabilities
cs.CL
Llama-Breeze2 (hereinafter referred to as Breeze2) is a suite of advanced multi-modal language models, available in 3B and 8B parameter configurations, specifically designed to enhance Traditional Chinese language representation. Building upon the Llama 3.2 model family, we continue the pre-training of Breeze2 on an extensive corpus to enhance the linguistic and cultural heritage of Traditional Chinese. In addition to language modeling capabilities, we significantly augment the models with function calling and vision understanding capabilities. At the time of this publication, as far as we are aware, absent reasoning-inducing prompts, Breeze2 are the strongest performing models in Traditional Chinese function calling and image understanding in its size class. The effectiveness of Breeze2 is benchmarked across various tasks, including Taiwan general knowledge, instruction-following, long context, function calling, and vision understanding. We are publicly releasing all Breeze2 models under the Llama 3.2 Community License. We also showcase the capabilities of the model running on mobile platform with a mobile application which we also open source.
2501.13923
Efficient Mitigation of Error Floors in Quantum Error Correction using Non-Binary Low-Density Parity-Check Codes
quant-ph cs.IT math.IT
In this paper, we propose an efficient method to reduce error floors in quantum error correction using non-binary low-density parity-check (LDPC) codes. We identify and classify cycle structures in the parity-check matrix where estimated noise becomes trapped, and develop tailored decoding methods for each cycle type. For Type-I cycles, we propose a method to make the difference between estimated and true noise degenerate. Type-II cycles are shown to be uncorrectable, while for Type-III cycles, we utilize the fact that cycles in non-binary LDPC codes do not necessarily correspond to codewords, allowing us to estimate the true noise. Our method significantly improves decoding performance and reduces error floors.
2501.13924
Towards Robust Multimodal Open-set Test-time Adaptation via Adaptive Entropy-aware Optimization
cs.CV cs.AI cs.LG
Test-time adaptation (TTA) has demonstrated significant potential in addressing distribution shifts between training and testing data. Open-set test-time adaptation (OSTTA) aims to adapt a source pre-trained model online to an unlabeled target domain that contains unknown classes. This task becomes more challenging when multiple modalities are involved. Existing methods have primarily focused on unimodal OSTTA, often filtering out low-confidence samples without addressing the complexities of multimodal data. In this work, we present Adaptive Entropy-aware Optimization (AEO), a novel framework specifically designed to tackle Multimodal Open-set Test-time Adaptation (MM-OSTTA) for the first time. Our analysis shows that the entropy difference between known and unknown samples in the target domain strongly correlates with MM-OSTTA performance. To leverage this, we propose two key components: Unknown-aware Adaptive Entropy Optimization (UAE) and Adaptive Modality Prediction Discrepancy Optimization (AMP). These components enhance the ability of model to distinguish unknown class samples during online adaptation by amplifying the entropy difference between known and unknown samples. To thoroughly evaluate our proposed methods in the MM-OSTTA setting, we establish a new benchmark derived from existing datasets. This benchmark includes two downstream tasks and incorporates five modalities. Extensive experiments across various domain shift situations demonstrate the efficacy and versatility of the AEO framework. Additionally, we highlight the strong performance of AEO in long-term and continual MM-OSTTA settings, both of which are challenging and highly relevant to real-world applications. Our source code is available at https://github.com/donghao51/AEO.
2501.13925
GeoPixel: Pixel Grounding Large Multimodal Model in Remote Sensing
cs.CV
Recent advances in large multimodal models (LMMs) have recognized fine-grained grounding as an imperative factor of visual understanding and dialogue. However, the benefits of such representation in LMMs are limited to the natural image domain, and these models perform poorly for remote sensing (RS). The distinct overhead viewpoint, scale variation, and presence of small objects in high-resolution RS imagery present a unique challenge in region-level comprehension. Moreover, the development of the grounding conversation capability of LMMs within RS is hindered by the lack of granular, RS domain-specific grounded data. Addressing these limitations, we propose GeoPixel - the first end-to-end high resolution RS-LMM that supports pixel-level grounding. This capability allows fine-grained visual perception by generating interleaved masks in conversation. GeoPixel supports up to 4K HD resolution in any aspect ratio, ideal for high-precision RS image analysis. To support the grounded conversation generation (GCG) in RS imagery, we curate a visually grounded dataset GeoPixelD through a semi-automated pipeline that utilizes set-of-marks prompting and spatial priors tailored for RS data to methodically control the data generation process. GeoPixel demonstrates superior performance in pixel-level comprehension, surpassing existing LMMs in both single-target and multi-target segmentation tasks. Our methodological ablation studies validate the effectiveness of each component in the overall architecture. Our code and data will be publicly released.
2501.13926
Can We Generate Images with CoT? Let's Verify and Reinforce Image Generation Step by Step
cs.CV cs.AI cs.CL
Chain-of-Thought (CoT) reasoning has been extensively explored in large models to tackle complex understanding tasks. However, it still remains an open question whether such strategies can be applied to verifying and reinforcing image generation scenarios. In this paper, we provide the first comprehensive investigation of the potential of CoT reasoning to enhance autoregressive image generation. We focus on three techniques: scaling test-time computation for verification, aligning model preferences with Direct Preference Optimization (DPO), and integrating these techniques for complementary effects. Our results demonstrate that these approaches can be effectively adapted and combined to significantly improve image generation performance. Furthermore, given the pivotal role of reward models in our findings, we propose the Potential Assessment Reward Model (PARM) and PARM++, specialized for autoregressive image generation. PARM adaptively assesses each generation step through a potential assessment approach, merging the strengths of existing reward models, and PARM++ further introduces a reflection mechanism to self-correct the generated unsatisfactory image. Using our investigated reasoning strategies, we enhance a baseline model, Show-o, to achieve superior results, with a significant +24% improvement on the GenEval benchmark, surpassing Stable Diffusion 3 by +15%. We hope our study provides unique insights and paves a new path for integrating CoT reasoning with autoregressive image generation. Code and models are released at https://github.com/ZiyuGuo99/Image-Generation-CoT
2501.13927
CRPO: Confidence-Reward Driven Preference Optimization for Machine Translation
cs.CL cs.AI cs.CV
Large language models (LLMs) have shown great potential in natural language processing tasks, but their application to machine translation (MT) remains challenging due to pretraining on English-centric data and the complexity of reinforcement learning from human feedback (RLHF). Direct Preference Optimization (DPO) has emerged as a simpler and more efficient alternative, but its performance depends heavily on the quality of preference data. To address this, we propose Confidence-Reward driven Preference Optimization (CRPO), a novel method that combines reward scores with model confidence to improve data selection for fine-tuning. CRPO selects challenging sentence pairs where the model is uncertain or underperforms, leading to more effective learning. While primarily designed for LLMs, CRPO also generalizes to encoder-decoder models like NLLB, demonstrating its versatility. Empirical results show that CRPO outperforms existing methods such as RS-DPO, RSO and MBR score in both translation accuracy and data efficiency.
2501.13928
Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass
cs.CV cs.AI cs.GR cs.RO
Multi-view 3D reconstruction remains a core challenge in computer vision, particularly in applications requiring accurate and scalable representations across diverse perspectives. Current leading methods such as DUSt3R employ a fundamentally pairwise approach, processing images in pairs and necessitating costly global alignment procedures to reconstruct from multiple views. In this work, we propose Fast 3D Reconstruction (Fast3R), a novel multi-view generalization to DUSt3R that achieves efficient and scalable 3D reconstruction by processing many views in parallel. Fast3R's Transformer-based architecture forwards N images in a single forward pass, bypassing the need for iterative alignment. Through extensive experiments on camera pose estimation and 3D reconstruction, Fast3R demonstrates state-of-the-art performance, with significant improvements in inference speed and reduced error accumulation. These results establish Fast3R as a robust alternative for multi-view applications, offering enhanced scalability without compromising reconstruction accuracy.
2501.13935
Low rank matrix completion and realization of graphs: results and problems
math.HO cs.DM cs.LG math.CO math.GT
The Netflix problem (from machine learning) asks the following. Given a ratings matrix in which each entry $(i,j)$ represents the rating of movie $j$ by customer $i$, if customer $i$ has watched movie $j$, and is otherwise missing, we would like to predict the remaining entries in order to make good recommendations to customers on what to watch next. The remaining entries are predicted so as to minimize the {\it rank} of the completed matrix. In this survey we study a more general problem, in which instead of knowing specific matrix elements, we know linear relations on such elements. We describe applications of these results to embeddings of graphs in surfaces (more precisely, embeddings with rotation systems, and embeddings modulo 2).
2501.13936
Evaluating Computational Accuracy of Large Language Models in Numerical Reasoning Tasks for Healthcare Applications
cs.AI cs.CL cs.LG
Large Language Models (LLMs) have emerged as transformative tools in the healthcare sector, demonstrating remarkable capabilities in natural language understanding and generation. However, their proficiency in numerical reasoning, particularly in high-stakes domains like in clinical applications, remains underexplored. Numerical reasoning is critical in healthcare applications, influencing patient outcomes, treatment planning, and resource allocation. This study investigates the computational accuracy of LLMs in numerical reasoning tasks within healthcare contexts. Using a curated dataset of 1,000 numerical problems, encompassing real-world scenarios such as dosage calculations and lab result interpretations, the performance of a refined LLM based on the GPT-3 architecture was evaluated. The methodology includes prompt engineering, integration of fact-checking pipelines, and application of regularization techniques to enhance model accuracy and generalization. Key metrics such as precision, recall, and F1-score were utilized to assess the model's efficacy. The results indicate an overall accuracy of 84.10%, with improved performance in straightforward numerical tasks and challenges in multi-step reasoning. The integration of a fact-checking pipeline improved accuracy by 11%, underscoring the importance of validation mechanisms. This research highlights the potential of LLMs in healthcare numerical reasoning and identifies avenues for further refinement to support critical decision-making in clinical environments. The findings aim to contribute to the development of reliable, interpretable, and contextually relevant AI tools for healthcare.
2501.13941
GaussMark: A Practical Approach for Structural Watermarking of Language Models
cs.CR cs.AI cs.CL cs.LG
Recent advances in Large Language Models (LLMs) have led to significant improvements in natural language processing tasks, but their ability to generate human-quality text raises significant ethical and operational concerns in settings where it is important to recognize whether or not a given text was generated by a human. Thus, recent work has focused on developing techniques for watermarking LLM-generated text, i.e., introducing an almost imperceptible signal that allows a provider equipped with a secret key to determine if given text was generated by their model. Current watermarking techniques are often not practical due to concerns with generation latency, detection time, degradation in text quality, or robustness. Many of these drawbacks come from the focus on token-level watermarking, which ignores the inherent structure of text. In this work, we introduce a new scheme, GaussMark, that is simple and efficient to implement, has formal statistical guarantees on its efficacy, comes at no cost in generation latency, and embeds the watermark into the weights of the model itself, providing a structural watermark. Our approach is based on Gaussian independence testing and is motivated by recent empirical observations that minor additive corruptions to LLM weights can result in models of identical (or even improved) quality. We show that by adding a small amount of Gaussian noise to the weights of a given LLM, we can watermark the model in a way that is statistically detectable by a provider who retains the secret key. We provide formal statistical bounds on the validity and power of our procedure. Through an extensive suite of experiments, we demonstrate that GaussMark is reliable, efficient, and relatively robust to corruptions such as insertions, deletions, substitutions, and roundtrip translations and can be instantiated with essentially no loss in model quality.
2501.13942
Prompt-Based Monte Carlo Tree Search for Mitigating Hallucinations in Large Models
cs.AI
With the rapid development of large models in the field of artificial intelligence, how to enhance their application capabilities in handling complex problems in the field of scientific research remains a challenging problem to be solved. This study proposes an improved Monte Carlo Tree Search (MCTS) method based on prompt words. In the simulation search stage, it introduces dynamic adjustment of exploration parameters and adaptive selection strategies, which can better balance exploration and exploitation, thereby reducing the hallucination phenomenon. This paper takes the four subsets of the SciEval dataset as the test objects, and compares the Glm-4-flash+Improved MCTS method with the methods of several existing models. The results show that the Improved MCTS method performs better, providing new ideas and methods for the application of large models in the field of scientific research.
2501.13943
Language Representation Favored Zero-Shot Cross-Domain Cognitive Diagnosis
cs.CL cs.AI cs.CY cs.LG
Cognitive diagnosis aims to infer students' mastery levels based on their historical response logs. However, existing cognitive diagnosis models (CDMs), which rely on ID embeddings, often have to train specific models on specific domains. This limitation may hinder their directly practical application in various target domains, such as different subjects (e.g., Math, English and Physics) or different education platforms (e.g., ASSISTments, Junyi Academy and Khan Academy). To address this issue, this paper proposes the language representation favored zero-shot cross-domain cognitive diagnosis (LRCD). Specifically, LRCD first analyzes the behavior patterns of students, exercises and concepts in different domains, and then describes the profiles of students, exercises and concepts using textual descriptions. Via recent advanced text-embedding modules, these profiles can be transformed to vectors in the unified language space. Moreover, to address the discrepancy between the language space and the cognitive diagnosis space, we propose language-cognitive mappers in LRCD to learn the mapping from the former to the latter. Then, these profiles can be easily and efficiently integrated and trained with existing CDMs. Extensive experiments show that training LRCD on real-world datasets can achieve commendable zero-shot performance across different target domains, and in some cases, it can even achieve competitive performance with some classic CDMs trained on the full response data on target domains. Notably, we surprisingly find that LRCD can also provide interesting insights into the differences between various subjects (such as humanities and sciences) and sources (such as primary and secondary education).
2501.13944
Fanar: An Arabic-Centric Multimodal Generative AI Platform
cs.CL cs.AI
We present Fanar, a platform for Arabic-centric multimodal generative AI systems, that supports language, speech and image generation tasks. At the heart of Fanar are Fanar Star and Fanar Prime, two highly capable Arabic Large Language Models (LLMs) that are best in the class on well established benchmarks for similar sized models. Fanar Star is a 7B (billion) parameter model that was trained from scratch on nearly 1 trillion clean and deduplicated Arabic, English and Code tokens. Fanar Prime is a 9B parameter model continually trained on the Gemma-2 9B base model on the same 1 trillion token set. Both models are concurrently deployed and designed to address different types of prompts transparently routed through a custom-built orchestrator. The Fanar platform provides many other capabilities including a customized Islamic Retrieval Augmented Generation (RAG) system for handling religious prompts, a Recency RAG for summarizing information about current or recent events that have occurred after the pre-training data cut-off date. The platform provides additional cognitive capabilities including in-house bilingual speech recognition that supports multiple Arabic dialects, voice and image generation that is fine-tuned to better reflect regional characteristics. Finally, Fanar provides an attribution service that can be used to verify the authenticity of fact based generated content. The design, development, and implementation of Fanar was entirely undertaken at Hamad Bin Khalifa University's Qatar Computing Research Institute (QCRI) and was sponsored by Qatar's Ministry of Communications and Information Technology to enable sovereign AI technology development.
2501.13945
Self-Explanation in Social AI Agents
cs.CL cs.AI cs.CY
Social AI agents interact with members of a community, thereby changing the behavior of the community. For example, in online learning, an AI social assistant may connect learners and thereby enhance social interaction. These social AI assistants too need to explain themselves in order to enhance transparency and trust with the learners. We present a method of self-explanation that uses introspection over a self-model of an AI social assistant. The self-model is captured as a functional model that specifies how the methods of the agent use knowledge to achieve its tasks. The process of generating self-explanations uses Chain of Thought to reflect on the self-model and ChatGPT to provide explanations about its functioning. We evaluate the self-explanation of the AI social assistant for completeness and correctness. We also report on its deployment in a live class.
2501.13946
Hallucination Mitigation using Agentic AI Natural Language-Based Frameworks
cs.CL cs.AI cs.MA
Hallucinations remain a significant challenge in current Generative AI models, undermining trust in AI systems and their reliability. This study investigates how orchestrating multiple specialized Artificial Intelligent Agents can help mitigate such hallucinations, with a focus on systems leveraging Natural Language Processing (NLP) to facilitate seamless agent interactions. To achieve this, we design a pipeline that introduces over three hundred prompts, purposefully crafted to induce hallucinations, into a front-end agent. The outputs are then systematically reviewed and refined by second- and third-level agents, each employing distinct large language models and tailored strategies to detect unverified claims, incorporate explicit disclaimers, and clarify speculative content. Additionally, we introduce a set of novel Key Performance Indicators (KPIs) specifically designed to evaluate hallucination score levels. A dedicated fourth-level AI agent is employed to evaluate these KPIs, providing detailed assessments and ensuring accurate quantification of shifts in hallucination-related behaviors. A core component of this investigation is the use of the OVON (Open Voice Network) framework, which relies on universal NLP-based interfaces to transfer contextual information among agents. Through structured JSON messages, each agent communicates its assessment of the hallucination likelihood and the reasons underlying questionable content, thereby enabling the subsequent stage to refine the text without losing context. The results demonstrate that employing multiple specialized agents capable of interoperating with each other through NLP-based agentic frameworks can yield promising outcomes in hallucination mitigation, ultimately bolstering trust within the AI community.
2501.13947
A Comprehensive Survey on Integrating Large Language Models with Knowledge-Based Methods
cs.CL cs.AI
The rapid development of artificial intelligence has brought about substantial advancements in the field. One promising direction is the integration of Large Language Models (LLMs) with structured knowledge-based systems. This approach aims to enhance AI capabilities by combining the generative language understanding of LLMs with the precise knowledge representation of structured systems. This survey explores the synergy between LLMs and knowledge bases, focusing on real-world applications and addressing associated technical, operational, and ethical challenges. Through a comprehensive literature review, the study identifies critical issues and evaluates existing solutions. The paper highlights the benefits of integrating generative AI with knowledge bases, including improved data contextualization, enhanced model accuracy, and better utilization of knowledge resources. The findings provide a detailed overview of the current state of research, identify key gaps, and offer actionable recommendations. These insights contribute to advancing AI technologies and support their practical deployment across various sectors.
2501.13948
Longitudinal Abuse and Sentiment Analysis of Hollywood Movie Dialogues using LLMs
cs.CL cs.AI
Over the past decades, there has been an increasing concern about the prevalence of abusive and violent content in Hollywood movies. This study uses Large Language Models (LLMs) to explore the longitudinal abuse and sentiment analysis of Hollywood Oscar and blockbuster movie dialogues from 1950 to 2024. By employing fine-tuned LLMs, we analyze subtitles for over a thousand movies categorised into four genres to examine the trends and shifts in emotional and abusive content over the past seven decades. Our findings reveal significant temporal changes in movie dialogues, which reflect broader social and cultural influences. Overall, the emotional tendencies in the films are diverse, and the detection of abusive content also exhibits significant fluctuations. The results show a gradual rise in abusive content in recent decades, reflecting social norms and regulatory policy changes. Genres such as thrillers still present a higher frequency of abusive content that emphasises the ongoing narrative role of violence and conflict. At the same time, underlying positive emotions such as humour and optimism remain prevalent in most of the movies. Furthermore, the gradual increase of abusive content in movie dialogues has been significant over the last two decades, where Oscar-nominated movies overtook the top ten blockbusters.
2501.13949
Can OpenAI o1 Reason Well in Ophthalmology? A 6,990-Question Head-to-Head Evaluation Study
cs.CL cs.AI
Question: What is the performance and reasoning ability of OpenAI o1 compared to other large language models in addressing ophthalmology-specific questions? Findings: This study evaluated OpenAI o1 and five LLMs using 6,990 ophthalmological questions from MedMCQA. O1 achieved the highest accuracy (0.88) and macro-F1 score but ranked third in reasoning capabilities based on text-generation metrics. Across subtopics, o1 ranked first in ``Lens'' and ``Glaucoma'' but second to GPT-4o in ``Corneal and External Diseases'', ``Vitreous and Retina'' and ``Oculoplastic and Orbital Diseases''. Subgroup analyses showed o1 performed better on queries with longer ground truth explanations. Meaning: O1's reasoning enhancements may not fully extend to ophthalmology, underscoring the need for domain-specific refinements to optimize performance in specialized fields like ophthalmology.
2501.13950
DEFEND: A Large-scale 1M Dataset and Foundation Model for Tobacco Addiction Prevention
cs.CV
While tobacco advertising innovates at unprecedented speed, traditional surveillance methods remain frozen in time, especially in the context of social media. The lack of large-scale, comprehensive datasets and sophisticated monitoring systems has created a widening gap between industry advancement and public health oversight. This paper addresses this critical challenge by introducing Tobacco-1M, a comprehensive dataset of one million tobacco product images with hierarchical labels spanning 75 product categories, and DEFEND, a novel foundation model for tobacco product understanding. Our approach integrates a Feature Enhancement Module for rich multimodal representation learning, a Local-Global Visual Coherence mechanism for detailed feature discrimination, and an Enhanced Image-Text Alignment strategy for precise product characterization. Experimental results demonstrate DEFEND's superior performance, achieving 83.1% accuracy in product classification and 73.8% in visual question-answering tasks, outperforming existing methods by significant margins. Moreover, the model exhibits robust zero-shot learning capabilities with 45.6% accuracy on novel product categories. This work provides regulatory bodies and public health researchers with powerful tools for monitoring emerging tobacco products and marketing strategies, potentially revolutionizing approaches to tobacco control and public health surveillance.
2501.13951
A Layered Multi-Expert Framework for Long-Context Mental Health Assessments
cs.CL cs.AI
Long-form mental health assessments pose unique challenges for large language models (LLMs), which often exhibit hallucinations or inconsistent reasoning when handling extended, domain-specific contexts. We introduce Stacked Multi-Model Reasoning (SMMR), a layered framework that leverages multiple LLMs and specialized smaller models as coequal 'experts'. Early layers isolate short, discrete subtasks, while later layers integrate and refine these partial outputs through more advanced long-context models. We evaluate SMMR on the DAIC-WOZ depression-screening dataset and 48 curated case studies with psychiatric diagnoses, demonstrating consistent improvements over single-model baselines in terms of accuracy, F1-score, and PHQ-8 error reduction. By harnessing diverse 'second opinions', SMMR mitigates hallucinations, captures subtle clinical nuances, and enhances reliability in high-stakes mental health assessments. Our findings underscore the value of multi-expert frameworks for more trustworthy AI-driven screening.
2501.13952
The Dual-use Dilemma in LLMs: Do Empowering Ethical Capacities Make a Degraded Utility?
cs.CL cs.AI
Recent years have witnessed extensive efforts to enhance Large Language Models (LLMs) across various domains, alongside growing attention to their ethical implications. However, a critical challenge remains largely overlooked: LLMs must balance between rejecting harmful requests for safety and accommodating legitimate ones for utility. This paper presents a Direct Preference Optimization (DPO) based alignment framework that achieves better overall performance by addressing this ethical-utility trade-off, using chemical domain applications as a proof-of-concept. Our alignment pipeline starts with a GPT-assisted three-phase data generation scheme, in which we create LibraChemQA, a chemical question-answering dataset comprising 31.6k triplet instances. By incorporating an innovative balanced seed in the data generation process, our framework systematically considers both legitimate and illegitimate requests. The framework also introduces a rephrasing mechanism for efficient data augmentation that enhances the model's chemical comprehension. We further develop a novel hybrid evaluation scheme with LLM judges for precise assessment of both safety and utility. Experimental results demonstrate our model's substantial improvements in overall performance where both safety and utility are considered - our resulting model, LibraChem, outperforms leading LLMs including Claude-3, GPT-4o, and LLaMA-3 by margins of 13.44%, 7.16%, and 7.10% respectively on our released benchmark.
2501.13953
Redundancy Principles for MLLMs Benchmarks
cs.CL cs.AI
With the rapid iteration of Multi-modality Large Language Models (MLLMs) and the evolving demands of the field, the number of benchmarks produced annually has surged into the hundreds. The rapid growth has inevitably led to significant redundancy among benchmarks. Therefore, it is crucial to take a step back and critically assess the current state of redundancy and propose targeted principles for constructing effective MLLM benchmarks. In this paper, we focus on redundancy from three key perspectives: 1) Redundancy of benchmark capability dimensions, 2) Redundancy in the number of test questions, and 3) Cross-benchmark redundancy within specific domains. Through the comprehensive analysis over hundreds of MLLMs' performance across more than 20 benchmarks, we aim to quantitatively measure the level of redundancy lies in existing MLLM evaluations, provide valuable insights to guide the future development of MLLM benchmarks, and offer strategies to refine and address redundancy issues effectively.