Showing 1–50 of 152 results for author: Zhai, X

Multi-Stage Prompt Inference Attacks on Enterprise LLM Systems

Authors: Andrii Balashov, Olena Ponomarova, Xiaohua Zhai

Abstract: Large Language Models (LLMs) deployed in enterprise settings (e.g., as Microsoft 365 Copilot) face novel security challenges. One critical threat is prompt inference attacks: adversaries chain together seemingly benign prompts to gradually extract confidential data. In this paper, we present a comprehensive study of multi-stage prompt inference attacks in an enterprise LLM context. We simulate rea… ▽ More Large Language Models (LLMs) deployed in enterprise settings (e.g., as Microsoft 365 Copilot) face novel security challenges. One critical threat is prompt inference attacks: adversaries chain together seemingly benign prompts to gradually extract confidential data. In this paper, we present a comprehensive study of multi-stage prompt inference attacks in an enterprise LLM context. We simulate realistic attack scenarios where an attacker uses mild-mannered queries and indirect prompt injections to exploit an LLM integrated with private corporate data. We develop a formal threat model for these multi-turn inference attacks and analyze them using probability theory, optimization frameworks, and information-theoretic leakage bounds. The attacks are shown to reliably exfiltrate sensitive information from the LLM's context (e.g., internal SharePoint documents or emails), even when standard safety measures are in place. We propose and evaluate defenses to counter such attacks, including statistical anomaly detection, fine-grained access control, prompt sanitization techniques, and architectural modifications to LLM deployment. Each defense is supported by mathematical analysis or experimental simulation. For example, we derive bounds on information leakage under differential privacy-based training and demonstrate an anomaly detection method that flags multi-turn attacks with high AUC. We also introduce an approach called "spotlighting" that uses input transformations to isolate untrusted prompt content, reducing attack success by an order of magnitude. Finally, we provide a formal proof of concept and empirical validation for a combined defense-in-depth strategy. Our work highlights that securing LLMs in enterprise settings requires moving beyond single-turn prompt filtering toward a holistic, multi-stage perspective on both attacks and defenses. △ Less

Submitted 21 July, 2025; originally announced July 2025.

Comments: 26 pages

SeqCSIST: Sequential Closely-Spaced Infrared Small Target Unmixing

Authors: Ximeng Zhai, Bohan Xu, Yaohong Chen, Hao Wang, Kehua Guo, Yimian Dai

Abstract: Due to the limitation of the optical lens focal length and the resolution of the infrared detector, distant Closely-Spaced Infrared Small Target (CSIST) groups typically appear as mixing spots in the infrared image. In this paper, we propose a novel task, Sequential CSIST Unmixing, namely detecting all targets in the form of sub-pixel localization from a highly dense CSIST group. However, achievin… ▽ More Due to the limitation of the optical lens focal length and the resolution of the infrared detector, distant Closely-Spaced Infrared Small Target (CSIST) groups typically appear as mixing spots in the infrared image. In this paper, we propose a novel task, Sequential CSIST Unmixing, namely detecting all targets in the form of sub-pixel localization from a highly dense CSIST group. However, achieving such precise detection is an extremely difficult challenge. In addition, the lack of high-quality public datasets has also restricted the research progress. To this end, firstly, we contribute an open-source ecosystem, including SeqCSIST, a sequential benchmark dataset, and a toolkit that provides objective evaluation metrics for this special task, along with the implementation of 23 relevant methods. Furthermore, we propose the Deformable Refinement Network (DeRefNet), a model-driven deep learning framework that introduces a Temporal Deformable Feature Alignment (TDFA) module enabling adaptive inter-frame information aggregation. To the best of our knowledge, this work is the first endeavor to address the CSIST Unmixing task within a multi-frame paradigm. Experiments on the SeqCSIST dataset demonstrate that our method outperforms the state-of-the-art approaches with mean Average Precision (mAP) metric improved by 5.3\%. Our dataset and toolkit are available from https://github.com/GrokCV/SeqCSIST. △ Less

Submitted 13 July, 2025; originally announced July 2025.

Comments: Accepted by TGRS

Towards Text-free Graph Foundation Models: Rethinking Multi-Domain Graph Contrastive Learning

Authors: Zihao Zhao, Xinlong Zhai, Jinyu Yang, Chuan Shi

Abstract: Foundation models have achieved great success in natural language processing (NLP) and computer vision (CV). Their success largely stems from the ability to integrate multi-domain knowledge in pre-training and transfer it to target domains. Considering graph data, especially graphs without textual features, is ubiquitous in real-world applications such as social networks and recommendation systems… ▽ More Foundation models have achieved great success in natural language processing (NLP) and computer vision (CV). Their success largely stems from the ability to integrate multi-domain knowledge in pre-training and transfer it to target domains. Considering graph data, especially graphs without textual features, is ubiquitous in real-world applications such as social networks and recommendation systems, some researchers have attempted to extend this paradigm to the graph field, aiming to construct graph foundation models. However, unlike CV and NLP, there are huge gaps among the semantics and properties of graphs in different domains, while current works still adopt traditional contrastive pre-training strategies designed in the single-domain scenario, which regard contrastive samples from different domains as equivalent. From experimental investigations, we discovered that inherent domain-specific differences prevent these strategies from effectively absorbing knowledge from different domains to generate informative representations. In this paper, we propose a novel multi-domain pre-training and cross-domain transfer framework, namely MDGCL.In the pre-training stage, we design a contrastive learning strategy to substantially recognize and capture domain differences, and introduce domain tokens to encode domain-level global information. In the downstream stage, we introduce a domain attention mechanism to enable fine-grained domain knowledge transfer. Extensive experiments on five benchmark datasets have demonstrated that our method outperforms state-of-the-art significantly, with the maximum improvement of 19.33\% on accuracy and 19.13\% on Macro-F1 score. △ Less

Submitted 25 June, 2025; originally announced June 2025.

Comments: 16 pages, 5 figures

Using Large Language Models to Assess Teachers' Pedagogical Content Knowledge

Authors: Yaxuan Yang, Shiyu Wang, Xiaoming Zhai

Abstract: Assessing teachers' pedagogical content knowledge (PCK) through performance-based tasks is both time and effort-consuming. While large language models (LLMs) offer new opportunities for efficient automatic scoring, little is known about whether LLMs introduce construct-irrelevant variance (CIV) in ways similar to or different from traditional machine learning (ML) and human raters. This study exam… ▽ More Assessing teachers' pedagogical content knowledge (PCK) through performance-based tasks is both time and effort-consuming. While large language models (LLMs) offer new opportunities for efficient automatic scoring, little is known about whether LLMs introduce construct-irrelevant variance (CIV) in ways similar to or different from traditional machine learning (ML) and human raters. This study examines three sources of CIV -- scenario variability, rater severity, and rater sensitivity to scenario -- in the context of video-based constructed-response tasks targeting two PCK sub-constructs: analyzing student thinking and evaluating teacher responsiveness. Using generalized linear mixed models (GLMMs), we compared variance components and rater-level scoring patterns across three scoring sources: human raters, supervised ML, and LLM. Results indicate that scenario-level variance was minimal across tasks, while rater-related factors contributed substantially to CIV, especially in the more interpretive Task II. The ML model was the most severe and least sensitive rater, whereas the LLM was the most lenient. These findings suggest that the LLM contributes to scoring efficiency while also introducing CIV as human raters do, yet with varying levels of contribution compared to supervised ML. Implications for rater training, automated scoring design, and future research on model interpretability are discussed. △ Less

Submitted 25 May, 2025; originally announced May 2025.

Fractal Graph Contrastive Learning

Authors: Nero Z. Li, Xuehao Zhai, Zhichao Shi, Boshen Shi, Xuhui Jiang

Abstract: While Graph Contrastive Learning (GCL) has attracted considerable attention in the field of graph self-supervised learning, its performance heavily relies on data augmentations that are expected to generate semantically consistent positive pairs. Existing strategies typically resort to random perturbations or local structure preservation, yet lack explicit control over global structural consistenc… ▽ More While Graph Contrastive Learning (GCL) has attracted considerable attention in the field of graph self-supervised learning, its performance heavily relies on data augmentations that are expected to generate semantically consistent positive pairs. Existing strategies typically resort to random perturbations or local structure preservation, yet lack explicit control over global structural consistency between augmented views. To address this limitation, we propose Fractal Graph Contrastive Learning (FractalGCL), a theory-driven framework that leverages fractal self-similarity to enforce global topological coherence. FractalGCL introduces two key innovations: a renormalisation-based augmentation that generates structurally aligned positive views via box coverings; and a fractal-dimension-aware contrastive loss that aligns graph embeddings according to their fractal dimensions. While combining the two innovations markedly boosts graph-representation quality, it also adds non-trivial computational overhead. To mitigate the computational overhead of fractal dimension estimation, we derive a one-shot estimator by proving that the dimension discrepancy between original and renormalised graphs converges weakly to a centred Gaussian distribution. This theoretical insight enables a reduction in dimension computation cost by an order of magnitude, cutting overall training time by approximately 61%. The experiments show that FractalGCL not only delivers state-of-the-art results on standard benchmarks but also outperforms traditional baselines on traffic networks by an average margin of about remarkably 7%. Codes are available at (https://anonymous.4open.science/r/FractalGCL-0511). △ Less

Submitted 22 May, 2025; v1 submitted 16 May, 2025; originally announced May 2025.

Artificial Intelligence Bias on English Language Learners in Automatic Scoring

Authors: Shuchen Guo, Yun Wang, Jichao Yu, Xuansheng Wu, Bilgehan Ayik, Field M. Watts, Ehsan Latif, Ninghao Liu, Lei Liu, Xiaoming Zhai

Abstract: This study investigated potential scoring biases and disparities toward English Language Learners (ELLs) when using automatic scoring systems for middle school students' written responses to science assessments. We specifically focus on examining how unbalanced training data with ELLs contributes to scoring bias and disparities. We fine-tuned BERT with four datasets: responses from (1) ELLs, (2) n… ▽ More This study investigated potential scoring biases and disparities toward English Language Learners (ELLs) when using automatic scoring systems for middle school students' written responses to science assessments. We specifically focus on examining how unbalanced training data with ELLs contributes to scoring bias and disparities. We fine-tuned BERT with four datasets: responses from (1) ELLs, (2) non-ELLs, (3) a mixed dataset reflecting the real-world proportion of ELLs and non-ELLs (unbalanced), and (4) a balanced mixed dataset with equal representation of both groups. The study analyzed 21 assessment items: 10 items with about 30,000 ELL responses, five items with about 1,000 ELL responses, and six items with about 200 ELL responses. Scoring accuracy (Acc) was calculated and compared to identify bias using Friedman tests. We measured the Mean Score Gaps (MSGs) between ELLs and non-ELLs and then calculated the differences in MSGs generated through both the human and AI models to identify the scoring disparities. We found that no AI bias and distorted disparities between ELLs and non-ELLs were found when the training dataset was large enough (ELL = 30,000 and ELL = 1,000), but concerns could exist if the sample size is limited (ELL = 200). △ Less

Submitted 19 May, 2025; v1 submitted 15 May, 2025; originally announced May 2025.

Understanding University Students' Use of Generative AI: The Roles of Demographics and Personality Traits

Authors: Newnew Deng, Edward Jiusi Liu, Xiaoming Zhai

Abstract: The use of generative AI (GAI) among university students is rapidly increasing, yet empirical research on students' GAI use and the factors influencing it remains limited. To address this gap, we surveyed 363 undergraduate and graduate students in the United States, examining their GAI usage and how it relates to demographic variables and personality traits based on the Big Five model (i.e., extra… ▽ More The use of generative AI (GAI) among university students is rapidly increasing, yet empirical research on students' GAI use and the factors influencing it remains limited. To address this gap, we surveyed 363 undergraduate and graduate students in the United States, examining their GAI usage and how it relates to demographic variables and personality traits based on the Big Five model (i.e., extraversion, agreeableness, conscientiousness, and emotional stability, and intellect/imagination). Our findings reveal: (a) Students in higher academic years are more inclined to use GAI and prefer it over traditional resources. (b) Non-native English speakers use and adopt GAI more readily than native speakers. (c) Compared to White, Asian students report higher GAI usage, perceive greater academic benefits, and express a stronger preference for it. Similarly, Black students report a more positive impact of GAI on their academic performance. Personality traits also play a significant role in shaping perceptions and usage of GAI. After controlling demographic factors, we found that personality still significantly predicts GAI use and attitudes: (a) Students with higher conscientiousness use GAI less. (b) Students who are higher in agreeableness perceive a less positive impact of GAI on academic performance and express more ethical concerns about using it for academic work. (c) Students with higher emotional stability report a more positive impact of GAI on learning and fewer concerns about its academic use. (d) Students with higher extraversion show a stronger preference for GAI over traditional resources. (e) Students with higher intellect/imagination tend to prefer traditional resources. These insights highlight the need for universities to provide personalized guidance to ensure students use GAI effectively, ethically, and equitably in their academic pursuits. △ Less

Submitted 19 May, 2025; v1 submitted 3 May, 2025; originally announced May 2025.

MDD-LLM: Towards Accuracy Large Language Models for Major Depressive Disorder Diagnosis

Authors: Yuyang Sha, Hongxin Pan, Wei Xu, Weiyu Meng, Gang Luo, Xinyu Du, Xiaobing Zhai, Henry H. Y. Tong, Caijuan Shi, Kefeng Li

Abstract: Major depressive disorder (MDD) impacts more than 300 million people worldwide, highlighting a significant public health issue. However, the uneven distribution of medical resources and the complexity of diagnostic methods have resulted in inadequate attention to this disorder in numerous countries and regions. This paper introduces a high-performance MDD diagnosis tool named MDD-LLM, an AI-driven… ▽ More Major depressive disorder (MDD) impacts more than 300 million people worldwide, highlighting a significant public health issue. However, the uneven distribution of medical resources and the complexity of diagnostic methods have resulted in inadequate attention to this disorder in numerous countries and regions. This paper introduces a high-performance MDD diagnosis tool named MDD-LLM, an AI-driven framework that utilizes fine-tuned large language models (LLMs) and extensive real-world samples to tackle challenges in MDD diagnosis. Therefore, we select 274,348 individual information from the UK Biobank cohort to train and evaluate the proposed method. Specifically, we select 274,348 individual records from the UK Biobank cohort and design a tabular data transformation method to create a large corpus for training and evaluating the proposed approach. To illustrate the advantages of MDD-LLM, we perform comprehensive experiments and provide several comparative analyses against existing model-based solutions across multiple evaluation metrics. Experimental results show that MDD-LLM (70B) achieves an accuracy of 0.8378 and an AUC of 0.8919 (95% CI: 0.8799 - 0.9040), significantly outperforming existing machine learning and deep learning frameworks for MDD diagnosis. Given the limited exploration of LLMs in MDD diagnosis, we examine numerous factors that may influence the performance of our proposed method, such as tabular data transformation techniques and different fine-tuning strategies. △ Less

Submitted 28 April, 2025; originally announced May 2025.

Knowledge Distillation and Dataset Distillation of Large Language Models: Emerging Trends, Challenges, and Future Directions

Authors: Luyang Fang, Xiaowei Yu, Jiazhang Cai, Yongkai Chen, Shushan Wu, Zhengliang Liu, Zhenyuan Yang, Haoran Lu, Xilin Gong, Yufang Liu, Terry Ma, Wei Ruan, Ali Abbasi, Jing Zhang, Tao Wang, Ehsan Latif, Wei Liu, Wei Zhang, Soheil Kolouri, Xiaoming Zhai, Dajiang Zhu, Wenxuan Zhong, Tianming Liu, Ping Ma

Abstract: The exponential growth of Large Language Models (LLMs) continues to highlight the need for efficient strategies to meet ever-expanding computational and data demands. This survey provides a comprehensive analysis of two complementary paradigms: Knowledge Distillation (KD) and Dataset Distillation (DD), both aimed at compressing LLMs while preserving their advanced reasoning capabilities and lingui… ▽ More The exponential growth of Large Language Models (LLMs) continues to highlight the need for efficient strategies to meet ever-expanding computational and data demands. This survey provides a comprehensive analysis of two complementary paradigms: Knowledge Distillation (KD) and Dataset Distillation (DD), both aimed at compressing LLMs while preserving their advanced reasoning capabilities and linguistic diversity. We first examine key methodologies in KD, such as task-specific alignment, rationale-based training, and multi-teacher frameworks, alongside DD techniques that synthesize compact, high-impact datasets through optimization-based gradient matching, latent space regularization, and generative synthesis. Building on these foundations, we explore how integrating KD and DD can produce more effective and scalable compression strategies. Together, these approaches address persistent challenges in model scalability, architectural heterogeneity, and the preservation of emergent LLM abilities. We further highlight applications across domains such as healthcare and education, where distillation enables efficient deployment without sacrificing performance. Despite substantial progress, open challenges remain in preserving emergent reasoning and linguistic diversity, enabling efficient adaptation to continually evolving teacher models and datasets, and establishing comprehensive evaluation protocols. By synthesizing methodological innovations, theoretical foundations, and practical insights, our survey charts a path toward sustainable, resource-efficient LLMs through the tighter integration of KD and DD principles. △ Less

Submitted 20 April, 2025; originally announced April 2025.

Zero-shot Autonomous Microscopy for Scalable and Intelligent Characterization of 2D Materials

Authors: Jingyun Yang, Ruoyan Avery Yin, Chi Jiang, Yuepeng Hu, Xiaokai Zhu, Xingjian Hu, Sutharsika Kumar, Xiao Wang, Xiaohua Zhai, Keran Rong, Yunyue Zhu, Tianyi Zhang, Zongyou Yin, Jing Kong, Neil Zhenqiang Gong, Zhichu Ren, Haozhe Wang

Abstract: Characterization of atomic-scale materials traditionally requires human experts with months to years of specialized training. Even for trained human operators, accurate and reliable characterization remains challenging when examining newly discovered materials such as two-dimensional (2D) structures. This bottleneck drives demand for fully autonomous experimentation systems capable of comprehendin… ▽ More Characterization of atomic-scale materials traditionally requires human experts with months to years of specialized training. Even for trained human operators, accurate and reliable characterization remains challenging when examining newly discovered materials such as two-dimensional (2D) structures. This bottleneck drives demand for fully autonomous experimentation systems capable of comprehending research objectives without requiring large training datasets. In this work, we present ATOMIC (Autonomous Technology for Optical Microscopy & Intelligent Characterization), an end-to-end framework that integrates foundation models to enable fully autonomous, zero-shot characterization of 2D materials. Our system integrates the vision foundation model (i.e., Segment Anything Model), large language models (i.e., ChatGPT), unsupervised clustering, and topological analysis to automate microscope control, sample scanning, image segmentation, and intelligent analysis through prompt engineering, eliminating the need for additional training. When analyzing typical MoS2 samples, our approach achieves 99.7% segmentation accuracy for single layer identification, which is equivalent to that of human experts. In addition, the integrated model is able to detect grain boundary slits that are challenging to identify with human eyes. Furthermore, the system retains robust accuracy despite variable conditions including defocus, color temperature fluctuations, and exposure variations. It is applicable to a broad spectrum of common 2D materials-including graphene, MoS2, WSe2, SnSe-regardless of whether they were fabricated via chemical vapor deposition or mechanical exfoliation. This work represents the implementation of foundation models to achieve autonomous analysis, establishing a scalable and data-efficient characterization paradigm that fundamentally transforms the approach to nanoscale materials research. △ Less

Submitted 14 April, 2025; originally announced April 2025.

Comments: 13 pages, 4 figures

Structured and sparse partial least squares coherence for multivariate cortico-muscular analysis

Authors: Jingyao Sun, Qilu Zhang, Di Ma, Tianyu Jia, Shijie Jia, Xiaoxue Zhai, Ruimou Xie, Ping-Ju Lin, Zhibin Li, Yu Pan, Linhong Ji, Chong Li

Abstract: Multivariate cortico-muscular analysis has recently emerged as a promising approach for evaluating the corticospinal neural pathway. However, current multivariate approaches encounter challenges such as high dimensionality and limited sample sizes, thus restricting their further applications. In this paper, we propose a structured and sparse partial least squares coherence algorithm (ssPLSC) to ex… ▽ More Multivariate cortico-muscular analysis has recently emerged as a promising approach for evaluating the corticospinal neural pathway. However, current multivariate approaches encounter challenges such as high dimensionality and limited sample sizes, thus restricting their further applications. In this paper, we propose a structured and sparse partial least squares coherence algorithm (ssPLSC) to extract shared latent space representations related to cortico-muscular interactions. Our approach leverages an embedded optimization framework by integrating a partial least squares (PLS)-based objective function, a sparsity constraint and a connectivity-based structured constraint, addressing the generalizability, interpretability and spatial structure. To solve the optimization problem, we develop an efficient alternating iterative algorithm within a unified framework and prove its convergence experimentally. Extensive experimental results from one synthetic and several real-world datasets have demonstrated that ssPLSC can achieve competitive or better performance over some representative multivariate cortico-muscular fusion methods, particularly in scenarios characterized by limited sample sizes and high noise levels. This study provides a novel multivariate fusion method for cortico-muscular analysis, offering a transformative tool for the evaluation of corticospinal pathway integrity in neurological disorders. △ Less

Submitted 14 June, 2025; v1 submitted 24 March, 2025; originally announced March 2025.

Comments: This work has been submitted to the IEEE for possible publication

A Multi-Agent Framework Integrating Large Language Models and Generative AI for Accelerated Metamaterial Design

Authors: Jie Tian, Martin Taylor Sobczak, Dhanush Patil, Jixin Hou, Lin Pang, Arunachalam Ramanathan, Libin Yang, Xianyan Chen, Yuval Golan, Xiaoming Zhai, Hongyue Sun, Kenan Song, Xianqiao Wang

Abstract: Metamaterials, renowned for their exceptional mechanical, electromagnetic, and thermal properties, hold transformative potential across diverse applications, yet their design remains constrained by labor-intensive trial-and-error methods and limited data interoperability. Here, we introduce CrossMatAgent -- a novel multi-agent framework that synergistically integrates large language models with st… ▽ More Metamaterials, renowned for their exceptional mechanical, electromagnetic, and thermal properties, hold transformative potential across diverse applications, yet their design remains constrained by labor-intensive trial-and-error methods and limited data interoperability. Here, we introduce CrossMatAgent -- a novel multi-agent framework that synergistically integrates large language models with state-of-the-art generative AI to revolutionize metamaterial design. By orchestrating a hierarchical team of agents -- each specializing in tasks such as pattern analysis, architectural synthesis, prompt engineering, and supervisory feedback -- our system leverages the multimodal reasoning of GPT-4o alongside the generative precision of DALL-E 3 and a fine-tuned Stable Diffusion XL model. This integrated approach automates data augmentation, enhances design fidelity, and produces simulation- and 3D printing-ready metamaterial patterns. Comprehensive evaluations, including CLIP-based alignment, SHAP interpretability analyses, and mechanical simulations under varied load conditions, demonstrate the framework's ability to generate diverse, reproducible, and application-ready designs. CrossMatAgent thus establishes a scalable, AI-driven paradigm that bridges the gap between conceptual innovation and practical realization, paving the way for accelerated metamaterial development. △ Less

Submitted 6 April, 2025; v1 submitted 25 March, 2025; originally announced March 2025.

Gemma 3 Technical Report

Authors: Gemma Team, Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, Louis Rouillard, Thomas Mesnard, Geoffrey Cideron, Jean-bastien Grill, Sabela Ramos, Edouard Yvinec, Michelle Casbon, Etienne Pot, Ivo Penchev, Gaël Liu, Francesco Visin, Kathleen Kenealy, Lucas Beyer, Xiaohai Zhai, Anton Tsitsulin , et al. (191 additional authors not shown)

Abstract: We introduce Gemma 3, a multimodal addition to the Gemma family of lightweight open models, ranging in scale from 1 to 27 billion parameters. This version introduces vision understanding abilities, a wider coverage of languages and longer context - at least 128K tokens. We also change the architecture of the model to reduce the KV-cache memory that tends to explode with long context. This is achie… ▽ More We introduce Gemma 3, a multimodal addition to the Gemma family of lightweight open models, ranging in scale from 1 to 27 billion parameters. This version introduces vision understanding abilities, a wider coverage of languages and longer context - at least 128K tokens. We also change the architecture of the model to reduce the KV-cache memory that tends to explode with long context. This is achieved by increasing the ratio of local to global attention layers, and keeping the span on local attention short. The Gemma 3 models are trained with distillation and achieve superior performance to Gemma 2 for both pre-trained and instruction finetuned versions. In particular, our novel post-training recipe significantly improves the math, chat, instruction-following and multilingual abilities, making Gemma3-4B-IT competitive with Gemma2-27B-IT and Gemma3-27B-IT comparable to Gemini-1.5-Pro across benchmarks. We release all our models to the community. △ Less

Submitted 25 March, 2025; originally announced March 2025.

Blend the Separated: Mixture of Synergistic Experts for Data-Scarcity Drug-Target Interaction Prediction

Authors: Xinlong Zhai, Chunchen Wang, Ruijia Wang, Jiazheng Kang, Shujie Li, Boyu Chen, Tengfei Ma, Zikai Zhou, Cheng Yang, Chuan Shi

Abstract: Drug-target interaction prediction (DTI) is essential in various applications including drug discovery and clinical application. There are two perspectives of input data widely used in DTI prediction: Intrinsic data represents how drugs or targets are constructed, and extrinsic data represents how drugs or targets are related to other biological entities. However, any of the two perspectives of in… ▽ More Drug-target interaction prediction (DTI) is essential in various applications including drug discovery and clinical application. There are two perspectives of input data widely used in DTI prediction: Intrinsic data represents how drugs or targets are constructed, and extrinsic data represents how drugs or targets are related to other biological entities. However, any of the two perspectives of input data can be scarce for some drugs or targets, especially for those unpopular or newly discovered. Furthermore, ground-truth labels for specific interaction types can also be scarce. Therefore, we propose the first method to tackle DTI prediction under input data and/or label scarcity. To make our model functional when only one perspective of input data is available, we design two separate experts to process intrinsic and extrinsic data respectively and fuse them adaptively according to different samples. Furthermore, to make the two perspectives complement each other and remedy label scarcity, two experts synergize with each other in a mutually supervised way to exploit the enormous unlabeled data. Extensive experiments on 3 real-world datasets under different extents of input data scarcity and/or label scarcity demonstrate our model outperforms states of the art significantly and steadily, with a maximum improvement of 53.53%. We also test our model without any data scarcity and it still outperforms current methods. △ Less

Submitted 19 March, 2025; originally announced March 2025.

Privacy-Preserved Automated Scoring using Federated Learning for Educational Research

Authors: Ehsan Latif, Xiaoming Zhai

Abstract: Data privacy remains a critical concern in educational research, requiring strict adherence to ethical standards and regulatory protocols. While traditional approaches rely on anonymization and centralized data collection, they often expose raw student data to security vulnerabilities and impose substantial logistical overhead. In this study, we propose a federated learning (FL) framework for auto… ▽ More Data privacy remains a critical concern in educational research, requiring strict adherence to ethical standards and regulatory protocols. While traditional approaches rely on anonymization and centralized data collection, they often expose raw student data to security vulnerabilities and impose substantial logistical overhead. In this study, we propose a federated learning (FL) framework for automated scoring of educational assessments that eliminates the need to share sensitive data across institutions. Our approach leverages parameter-efficient fine-tuning of large language models (LLMs) with Low-Rank Adaptation (LoRA), enabling each client (school) to train locally while sharing only optimized model updates. To address data heterogeneity, we implement an adaptive weighted aggregation strategy that considers both client performance and data volume. We benchmark our model against two state-of-the-art FL methods and a centralized learning baseline using NGSS-aligned multi-label science assessment data from nine middle schools. Results show that our model achieves the highest accuracy (94.5%) among FL approaches, and performs within 0.5-1.0 percentage points of the centralized model on these metrics. Additionally, it achieves comparable rubric-level scoring accuracy, with only a 1.3% difference in rubric match and a lower score deviation (MAE), highlighting its effectiveness in preserving both prediction quality and interpretability. △ Less

Submitted 8 May, 2025; v1 submitted 12 March, 2025; originally announced March 2025.

Comments: Accepted to AIED25

Efficient Multi-Task Inferencing: Model Merging with Gromov-Wasserstein Feature Alignment

Authors: Luyang Fang, Ehsan Latif, Haoran Lu, Yifan Zhou, Ping Ma, Xiaoming Zhai

Abstract: Automatic scoring of student responses enhances efficiency in education, but deploying a separate neural network for each task increases storage demands, maintenance efforts, and redundant computations. To address these challenges, this paper introduces the Gromov-Wasserstein Scoring Model Merging (GW-SMM) method, which merges models based on feature distribution similarities measured via the Grom… ▽ More Automatic scoring of student responses enhances efficiency in education, but deploying a separate neural network for each task increases storage demands, maintenance efforts, and redundant computations. To address these challenges, this paper introduces the Gromov-Wasserstein Scoring Model Merging (GW-SMM) method, which merges models based on feature distribution similarities measured via the Gromov-Wasserstein distance. Our approach begins by extracting features from student responses using individual models, capturing both item-specific context and unique learned representations. The Gromov-Wasserstein distance then quantifies the similarity between these feature distributions, identifying the most compatible models for merging. Models exhibiting the smallest pairwise distances, typically in pairs or trios, are merged by combining only the shared layers preceding the classification head. This strategy results in a unified feature extractor while preserving separate classification heads for item-specific scoring. We validated our approach against human expert knowledge and a GPT-o1-based merging method. GW-SMM consistently outperformed both, achieving a higher micro F1 score, macro F1 score, exact match accuracy, and per-label accuracy. The improvements in micro F1 and per-label accuracy were statistically significant compared to GPT-o1-based merging (p=0.04, p=0.01). Additionally, GW-SMM reduced storage requirements by half without compromising much accuracy, demonstrating its computational efficiency alongside reliable scoring performance. △ Less

Submitted 12 March, 2025; originally announced March 2025.

Comments: Submitted to AIED2025

Advancing Education through Tutoring Systems: A Systematic Literature Review

Authors: Vincent Liu, Ehsan Latif, Xiaoming Zhai

Abstract: This study systematically reviews the transformative role of Tutoring Systems, encompassing Intelligent Tutoring Systems (ITS) and Robot Tutoring Systems (RTS), in addressing global educational challenges through advanced technologies. As many students struggle with proficiency in core academic areas, Tutoring Systems emerge as promising solutions to bridge learning gaps by delivering personalized… ▽ More This study systematically reviews the transformative role of Tutoring Systems, encompassing Intelligent Tutoring Systems (ITS) and Robot Tutoring Systems (RTS), in addressing global educational challenges through advanced technologies. As many students struggle with proficiency in core academic areas, Tutoring Systems emerge as promising solutions to bridge learning gaps by delivering personalized and adaptive instruction. ITS leverages artificial intelligence (AI) models, such as Bayesian Knowledge Tracing and Large Language Models, to provide precise cognitive support, while RTS enhances social and emotional engagement through human-like interactions. This systematic review, adhering to the PRISMA framework, analyzed 86 representative studies. We evaluated the pedagogical and technological advancements, engagement strategies, and ethical considerations surrounding these systems. Based on these parameters, Latent Class Analysis was conducted and identified three distinct categories: computer-based ITS, robot-based RTS, and multimodal systems integrating various interaction modes. The findings reveal significant advancements in AI techniques that enhance adaptability, engagement, and learning outcomes. However, challenges such as ethical concerns, scalability issues, and gaps in cognitive adaptability persist. The study highlights the complementary strengths of ITS and RTS, proposing integrated hybrid solutions to maximize educational benefits. Future research should focus on bridging gaps in scalability, addressing ethical considerations comprehensively, and advancing AI models to support diverse educational needs. △ Less

Submitted 12 March, 2025; originally announced March 2025.

Comments: A comprehensive study on tutoring systems

Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders

Authors: Xuansheng Wu, Jiayi Yuan, Wenlin Yao, Xiaoming Zhai, Ninghao Liu

Abstract: Large language models (LLMs) excel at handling human queries, but they can occasionally generate flawed or unexpected responses. Understanding their internal states is crucial for understanding their successes, diagnosing their failures, and refining their capabilities. Although sparse autoencoders (SAEs) have shown promise for interpreting LLM internal representations, limited research has explor… ▽ More Large language models (LLMs) excel at handling human queries, but they can occasionally generate flawed or unexpected responses. Understanding their internal states is crucial for understanding their successes, diagnosing their failures, and refining their capabilities. Although sparse autoencoders (SAEs) have shown promise for interpreting LLM internal representations, limited research has explored how to better explain SAE features, i.e., understanding the semantic meaning of features learned by SAE. Our theoretical analysis reveals that existing explanation methods suffer from the frequency bias issue, where they emphasize linguistic patterns over semantic concepts, while the latter is more critical to steer LLM behaviors. To address this, we propose using a fixed vocabulary set for feature interpretations and designing a mutual information-based objective, aiming to better capture the semantic meaning behind these features. We further propose two runtime steering strategies that adjust the learned feature activations based on their corresponding explanations. Empirical results show that, compared to baselines, our method provides more discourse-level explanations and effectively steers LLM behaviors to defend against jailbreak attacks. These findings highlight the value of explanations for steering LLM behaviors in downstream applications. We will release our code and data once accepted. △ Less

Submitted 21 February, 2025; originally announced February 2025.

Comments: Pre-print. 20 pages, 5 figures

SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features

Authors: Michael Tschannen, Alexey Gritsenko, Xiao Wang, Muhammad Ferjad Naeem, Ibrahim Alabdulmohsin, Nikhil Parthasarathy, Talfan Evans, Lucas Beyer, Ye Xia, Basil Mustafa, Olivier Hénaff, Jeremiah Harmsen, Andreas Steiner, Xiaohua Zhai

Abstract: We introduce SigLIP 2, a family of new multilingual vision-language encoders that build on the success of the original SigLIP. In this second iteration, we extend the original image-text training objective with several prior, independently developed techniques into a unified recipe -- this includes captioning-based pretraining, self-supervised losses (self-distillation, masked prediction) and onli… ▽ More We introduce SigLIP 2, a family of new multilingual vision-language encoders that build on the success of the original SigLIP. In this second iteration, we extend the original image-text training objective with several prior, independently developed techniques into a unified recipe -- this includes captioning-based pretraining, self-supervised losses (self-distillation, masked prediction) and online data curation. With these changes, SigLIP 2 models outperform their SigLIP counterparts at all model scales in core capabilities, including zero-shot classification, image-text retrieval, and transfer performance when extracting visual representations for Vision-Language Models (VLMs). Furthermore, the new training recipe leads to significant improvements on localization and dense prediction tasks. We also train variants which support multiple resolutions and preserve the input's native aspect ratio. Finally, we train on a more diverse data-mixture that includes de-biasing techniques, leading to much better multilingual understanding and improved fairness. To allow users to trade off inference cost with performance, we release model checkpoints at four sizes: ViT-B (86M), L (303M), So400m (400M), and g (1B). △ Less

Submitted 20 February, 2025; originally announced February 2025.

Comments: Model checkpoints are available at https://github.com/google-research/big_vision/tree/main/big_vision/configs/proj/image_text/README_siglip2.md

Self-Regularization with Sparse Autoencoders for Controllable LLM-based Classification

Authors: Xuansheng Wu, Wenhao Yu, Xiaoming Zhai, Ninghao Liu

Abstract: Modern text classification methods heavily rely on contextual embeddings from large language models (LLMs). Compared to human-engineered features, these embeddings provide automatic and effective representations for classification model training. However, they also introduce a challenge: we lose the ability to manually remove unintended features, such as sensitive or task-irrelevant features, to g… ▽ More Modern text classification methods heavily rely on contextual embeddings from large language models (LLMs). Compared to human-engineered features, these embeddings provide automatic and effective representations for classification model training. However, they also introduce a challenge: we lose the ability to manually remove unintended features, such as sensitive or task-irrelevant features, to guarantee regulatory compliance or improve the generalizability of classification models. This limitation arises because LLM embeddings are opaque and difficult to interpret. In this paper, we propose a novel framework to identify and regularize unintended features in the LLM latent space. Specifically, we first pre-train a sparse autoencoder (SAE) to extract interpretable features from LLM latent spaces. To ensure the SAE can capture task-specific features, we further fine-tune it on task-specific datasets. In training the classification model, we propose a simple and effective regularizer, by minimizing the similarity between the classifier weights and the identified unintended feature, to remove the impact of these unintended features on classification. We evaluate the proposed framework on three real-world tasks, including toxic chat detection, reward modeling, and disease diagnosis. Results show that the proposed self-regularization framework can improve the classifier's generalizability by regularizing those features that are not semantically correlated to the task. This work pioneers controllable text classification on LLM latent spaces by leveraging interpreted features to address generalizability, fairness, and privacy challenges. The code and data are publicly available at https://github.com/JacksonWuxs/Controllable_LLM_Classifier. △ Less

Submitted 15 June, 2025; v1 submitted 19 February, 2025; originally announced February 2025.

Comments: Accepted by SIGKDD 2025

Scaling Pre-training to One Hundred Billion Data for Vision Language Models

Authors: Xiao Wang, Ibrahim Alabdulmohsin, Daniel Salz, Zhe Li, Keran Rong, Xiaohua Zhai

Abstract: We provide an empirical investigation of the potential of pre-training vision-language models on an unprecedented scale: 100 billion examples. We find that model performance tends to saturate at this scale on many common Western-centric classification and retrieval benchmarks, such as COCO Captions. Nevertheless, tasks of cultural diversity achieve more substantial gains from the 100-billion scale… ▽ More We provide an empirical investigation of the potential of pre-training vision-language models on an unprecedented scale: 100 billion examples. We find that model performance tends to saturate at this scale on many common Western-centric classification and retrieval benchmarks, such as COCO Captions. Nevertheless, tasks of cultural diversity achieve more substantial gains from the 100-billion scale web data, thanks to its coverage of long-tail concepts. Furthermore, we analyze the model's multilinguality and show gains in low-resource languages as well. In addition, we observe that reducing the size of the pretraining dataset via quality filters like using CLIP, typically used to enhance performance, may inadvertently reduce the cultural diversity represented even in large-scale datasets. Our results highlight that while traditional benchmarks may not benefit significantly from scaling noisy, raw web data to 100 billion examples, this data scale is vital for building truly inclusive multimodal systems. △ Less

Submitted 11 February, 2025; originally announced February 2025.

Recursive Inference Scaling: A Winning Path to Scalable Inference in Language and Multimodal Systems

Authors: Ibrahim Alabdulmohsin, Xiaohua Zhai

Abstract: Inspired by recent findings on the fractal geometry of language, we introduce Recursive INference Scaling (RINS) as a complementary, plug-in recipe for scaling inference time in language and multimodal systems. RINS is a particular form of recursive depth that significantly outperforms +55 other variants, including the recent "repeat-all-over" (RAO) strategy in Mobile LLM (Liu et al., 2024) and la… ▽ More Inspired by recent findings on the fractal geometry of language, we introduce Recursive INference Scaling (RINS) as a complementary, plug-in recipe for scaling inference time in language and multimodal systems. RINS is a particular form of recursive depth that significantly outperforms +55 other variants, including the recent "repeat-all-over" (RAO) strategy in Mobile LLM (Liu et al., 2024) and latent recurrent thinking (Geiping et al., 2025). Unlike prior works, we carry out our comparisons on a compute-matched regime, and demonstrate that for a fixed model size and training compute budget, RINS substantially improves language modeling performance. It also generalizes beyond pure language tasks, delivering gains in multimodal systems, including a +2% improvement in 0-shot ImageNet accuracy for SigLIP-B/16. Additionally, by deriving data scaling laws, we show that RINS improves both the asymptotic performance limits and the scaling exponents. More importantly, with light-weight (linear) adapters (comprising <1% of model parameters) and stochastic dropout, RINS offers a no-regret strategy, meaning that RINS-enabled pretraining improves performance in language modeling even when recursive depth is not applied at inference time. This corresponds to improving performance on a training compute-, parameter-, and inference-matched regime, suggesting its potential as a viable component of LLM pretraining! △ Less

Submitted 8 May, 2025; v1 submitted 11 February, 2025; originally announced February 2025.

360Brew: A Decoder-only Foundation Model for Personalized Ranking and Recommendation

Authors: Hamed Firooz, Maziar Sanjabi, Adrian Englhardt, Aman Gupta, Ben Levine, Dre Olgiati, Gungor Polatkan, Iuliia Melnychuk, Karthik Ramgopal, Kirill Talanine, Kutta Srinivasan, Luke Simon, Natesh Sivasubramoniapillai, Necip Fazil Ayan, Qingquan Song, Samira Sriram, Souvik Ghosh, Tao Song, Tejas Dharamsi, Vignesh Kothapalli, Xiaoling Zhai, Ya Xu, Yu Wang, Yun Dai

Abstract: Ranking and recommendation systems are the foundation for numerous online experiences, ranging from search results to personalized content delivery. These systems have evolved into complex, multilayered architectures that leverage vast datasets and often incorporate thousands of predictive models. The maintenance and enhancement of these models is a labor intensive process that requires extensive… ▽ More Ranking and recommendation systems are the foundation for numerous online experiences, ranging from search results to personalized content delivery. These systems have evolved into complex, multilayered architectures that leverage vast datasets and often incorporate thousands of predictive models. The maintenance and enhancement of these models is a labor intensive process that requires extensive feature engineering. This approach not only exacerbates technical debt but also hampers innovation in extending these systems to emerging problem domains. In this report, we present our research to address these challenges by utilizing a large foundation model with a textual interface for ranking and recommendation tasks. We illustrate several key advantages of our approach: (1) a single model can manage multiple predictive tasks involved in ranking and recommendation, (2) decoder models with textual interface due to their comprehension of reasoning capabilities, can generalize to new recommendation surfaces and out-of-domain problems, and (3) by employing natural language interfaces for task definitions and verbalizing member behaviors and their social connections, we eliminate the need for feature engineering and the maintenance of complex directed acyclic graphs of model dependencies. We introduce our research pre-production model, 360Brew V1.0, a 150B parameter, decoder-only model that has been trained and fine-tuned on LinkedIn's data and tasks. This model is capable of solving over 30 predictive tasks across various segments of the LinkedIn platform, achieving performance levels comparable to or exceeding those of current production systems based on offline metrics, without task-specific fine-tuning. Notably, each of these tasks is conventionally addressed by dedicated models that have been developed and maintained over multiple years by teams of a similar or larger size than our own. △ Less

Submitted 7 February, 2025; v1 submitted 27 January, 2025; originally announced January 2025.

Fine-tuning ChatGPT for Automatic Scoring of Written Scientific Explanations in Chinese

Authors: Jie Yang, Ehsan Latif, Yuze He, Xiaoming Zhai

Abstract: The development of explanations for scientific phenomena is essential in science assessment, but scoring student-written explanations remains challenging and resource-intensive. Large language models (LLMs) have shown promise in addressing this issue, particularly in alphabetic languages like English. However, their applicability to logographic languages is less explored. This study investigates t… ▽ More The development of explanations for scientific phenomena is essential in science assessment, but scoring student-written explanations remains challenging and resource-intensive. Large language models (LLMs) have shown promise in addressing this issue, particularly in alphabetic languages like English. However, their applicability to logographic languages is less explored. This study investigates the potential of fine-tuning ChatGPT, a leading LLM, to automatically score scientific explanations written in Chinese. Student responses to seven scientific explanation tasks were collected and automatically scored, with scoring accuracy examined in relation to reasoning complexity using the Kendall correlation. A qualitative analysis explored how linguistic features influenced scoring accuracy. The results show that domain-specific adaptation enables ChatGPT to score Chinese scientific explanations with accuracy. However, scoring accuracy correlates with reasoning complexity: a negative correlation for lower-level responses and a positive one for higher-level responses. The model overrates complex reasoning in low-level responses with intricate sentence structures and underrates high-level responses using concise causal reasoning. These correlations stem from linguistic features--simplicity and clarity enhance accuracy for lower-level responses, while comprehensiveness improves accuracy for higher-level ones. Simpler, shorter responses tend to score more accurately at lower levels, whereas longer, information-rich responses yield better accuracy at higher levels. These findings demonstrate the effectiveness of LLMs in automatic scoring within a Chinese context and emphasize the importance of linguistic features and reasoning complexity in fine-tuning scoring models for educational assessments. △ Less

Submitted 11 January, 2025; originally announced January 2025.

Is Your Autonomous Vehicle Safe? Understanding the Threat of Electromagnetic Signal Injection Attacks on Traffic Scene Perception

Authors: Wenhao Liao, Sineng Yan, Youqian Zhang, Xinwei Zhai, Yuanyuan Wang, Eugene Yujun Fu

Abstract: Autonomous vehicles rely on camera-based perception systems to comprehend their driving environment and make crucial decisions, thereby ensuring vehicles to steer safely. However, a significant threat known as Electromagnetic Signal Injection Attacks (ESIA) can distort the images captured by these cameras, leading to incorrect AI decisions and potentially compromising the safety of autonomous vehi… ▽ More Autonomous vehicles rely on camera-based perception systems to comprehend their driving environment and make crucial decisions, thereby ensuring vehicles to steer safely. However, a significant threat known as Electromagnetic Signal Injection Attacks (ESIA) can distort the images captured by these cameras, leading to incorrect AI decisions and potentially compromising the safety of autonomous vehicles. Despite the serious implications of ESIA, there is limited understanding of its impacts on the robustness of AI models across various and complex driving scenarios. To address this gap, our research analyzes the performance of different models under ESIA, revealing their vulnerabilities to the attacks. Moreover, due to the challenges in obtaining real-world attack data, we develop a novel ESIA simulation method and generate a simulated attack dataset for different driving scenarios. Our research provides a comprehensive simulation and evaluation framework, aiming to enhance the development of more robust AI models and secure intelligent systems, ultimately contributing to the advancement of safer and more reliable technology across various fields. △ Less

Submitted 9 January, 2025; originally announced January 2025.

Comments: To appear in AAAI 2025

Human-Centered Design for AI-based Automatically Generated Assessment Reports: A Systematic Review

Authors: Ehsan Latif, Ying Chen, Xiaoming Zhai, Yue Yin

Abstract: This paper provides a comprehensive review of the design and implementation of automatically generated assessment reports (AutoRs) for formative use in K-12 Science, Technology, Engineering, and Mathematics (STEM) classrooms. With the increasing adoption of technology-enhanced assessments, there is a critical need for human-computer interactive tools that efficiently support the interpretation and… ▽ More This paper provides a comprehensive review of the design and implementation of automatically generated assessment reports (AutoRs) for formative use in K-12 Science, Technology, Engineering, and Mathematics (STEM) classrooms. With the increasing adoption of technology-enhanced assessments, there is a critical need for human-computer interactive tools that efficiently support the interpretation and application of assessment data by teachers. AutoRs are designed to provide synthesized, interpretable, and actionable insights into students' performance, learning progress, and areas for improvement. Guided by cognitive load theory, this study emphasizes the importance of reducing teachers' cognitive demands through user-centered and intuitive designs. It highlights the potential of diverse information presentation formats such as text, visual aids, and plots and advanced functionalities such as live and interactive features to enhance usability. However, the findings also reveal that many existing AutoRs fail to fully utilize these approaches, leading to high initial cognitive demands and limited engagement. This paper proposes a conceptual framework to inform the design, implementation, and evaluation of AutoRs, balancing the trade-offs between usability and functionality. The framework aims to address challenges in engaging teachers with technology-enhanced assessment results, facilitating data-driven decision-making, and providing personalized feedback to improve the teaching and learning process. △ Less

Submitted 30 December, 2024; originally announced January 2025.

Comments: Submitted to ETRD

Efficient Multi-Task Inferencing with a Shared Backbone and Lightweight Task-Specific Adapters for Automatic Scoring

Authors: Ehsan Latif, Xiaoming Zhai

Abstract: The integration of Artificial Intelligence (AI) in education requires scalable and efficient frameworks that balance performance, adaptability, and cost. This paper addresses these needs by proposing a shared backbone model architecture enhanced with lightweight LoRA adapters for task-specific fine-tuning, targeting the automated scoring of student responses across 27 mutually exclusive tasks. By… ▽ More The integration of Artificial Intelligence (AI) in education requires scalable and efficient frameworks that balance performance, adaptability, and cost. This paper addresses these needs by proposing a shared backbone model architecture enhanced with lightweight LoRA adapters for task-specific fine-tuning, targeting the automated scoring of student responses across 27 mutually exclusive tasks. By achieving competitive performance (average QWK of 0.848 compared to 0.888 for fully fine-tuned models) while reducing GPU memory consumption by 60% and inference latency by 40%, the framework demonstrates significant efficiency gains. This approach aligns with the workshop's focus on improving language models for educational tasks, creating responsible innovations for cost-sensitive deployment, and supporting educators by streamlining assessment workflows. The findings underscore the potential of scalable AI to enhance learning outcomes while maintaining fairness and transparency in automated scoring systems. △ Less

Submitted 21 June, 2025; v1 submitted 30 December, 2024; originally announced December 2024.

Comments: Accepted by AAAI-iRAISE Workshop

Can OpenAI o1 outperform humans in higher-order cognitive thinking?

Authors: Ehsan Latif, Yifan Zhou, Shuchen Guo, Lehong Shi, Yizhu Gao, Matthew Nyaaba, Arne Bewerdorff, Xiantong Yang, Xiaoming Zhai

Abstract: This study evaluates the performance of OpenAI's o1-preview model in higher-order cognitive domains, including critical thinking, systematic thinking, computational thinking, data literacy, creative thinking, logical reasoning, and scientific reasoning. Using established benchmarks, we compared the o1-preview models's performance to human participants from diverse educational levels. o1-preview ac… ▽ More This study evaluates the performance of OpenAI's o1-preview model in higher-order cognitive domains, including critical thinking, systematic thinking, computational thinking, data literacy, creative thinking, logical reasoning, and scientific reasoning. Using established benchmarks, we compared the o1-preview models's performance to human participants from diverse educational levels. o1-preview achieved a mean score of 24.33 on the Ennis-Weir Critical Thinking Essay Test (EWCTET), surpassing undergraduate (13.8) and postgraduate (18.39) participants (z = 1.60 and 0.90, respectively). In systematic thinking, it scored 46.1, SD = 4.12 on the Lake Urmia Vignette, significantly outperforming the human mean (20.08, SD = 8.13, z = 3.20). For data literacy, o1-preview scored 8.60, SD = 0.70 on Merk et al.'s "Use Data" dimension, compared to the human post-test mean of 4.17, SD = 2.02 (z = 2.19). On creative thinking tasks, the model achieved originality scores of 2.98, SD = 0.73, higher than the human mean of 1.74 (z = 0.71). In logical reasoning (LogiQA), it outperformed humans with average 90%, SD = 10% accuracy versus 86%, SD = 6.5% (z = 0.62). For scientific reasoning, it achieved near-perfect performance (mean = 0.99, SD = 0.12) on the TOSLS,, exceeding the highest human scores of 0.85, SD = 0.13 (z = 1.78). While o1-preview excelled in structured tasks, it showed limitations in problem-solving and adaptive reasoning. These results demonstrate the potential of AI to complement education in structured assessments but highlight the need for ethical oversight and refinement for broader applications. △ Less

Submitted 7 December, 2024; originally announced December 2024.

Foundation Models for Low-Resource Language Education (Vision Paper)

Authors: Zhaojun Ding, Zhengliang Liu, Hanqi Jiang, Yizhu Gao, Xiaoming Zhai, Tianming Liu, Ninghao Liu

Abstract: Recent studies show that large language models (LLMs) are powerful tools for working with natural language, bringing advances in many areas of computational linguistics. However, these models face challenges when applied to low-resource languages due to limited training data and difficulty in understanding cultural nuances. Research is now focusing on multilingual models to improve LLM performance… ▽ More Recent studies show that large language models (LLMs) are powerful tools for working with natural language, bringing advances in many areas of computational linguistics. However, these models face challenges when applied to low-resource languages due to limited training data and difficulty in understanding cultural nuances. Research is now focusing on multilingual models to improve LLM performance for these languages. Education in these languages also struggles with a lack of resources and qualified teachers, particularly in underdeveloped regions. Here, LLMs can be transformative, supporting innovative methods like community-driven learning and digital platforms. This paper discusses how LLMs could enhance education for low-resource languages, emphasizing practical applications and benefits. △ Less

Submitted 5 December, 2024; originally announced December 2024.

PaliGemma 2: A Family of Versatile VLMs for Transfer

Authors: Andreas Steiner, André Susano Pinto, Michael Tschannen, Daniel Keysers, Xiao Wang, Yonatan Bitton, Alexey Gritsenko, Matthias Minderer, Anthony Sherbondy, Shangbang Long, Siyang Qin, Reeve Ingle, Emanuele Bugliarello, Sahar Kazemzadeh, Thomas Mesnard, Ibrahim Alabdulmohsin, Lucas Beyer, Xiaohua Zhai

Abstract: PaliGemma 2 is an upgrade of the PaliGemma open Vision-Language Model (VLM) based on the Gemma 2 family of language models. We combine the SigLIP-So400m vision encoder that was also used by PaliGemma with the whole range of Gemma 2 models, from the 2B one all the way up to the 27B model. We train these models at three resolutions (224px, 448px, and 896px) in multiple stages to equip them with broa… ▽ More PaliGemma 2 is an upgrade of the PaliGemma open Vision-Language Model (VLM) based on the Gemma 2 family of language models. We combine the SigLIP-So400m vision encoder that was also used by PaliGemma with the whole range of Gemma 2 models, from the 2B one all the way up to the 27B model. We train these models at three resolutions (224px, 448px, and 896px) in multiple stages to equip them with broad knowledge for transfer via fine-tuning. The resulting family of base models covering different model sizes and resolutions allows us to investigate factors impacting transfer performance (such as learning rate) and to analyze the interplay between the type of task, model size, and resolution. We further increase the number and breadth of transfer tasks beyond the scope of PaliGemma including different OCR-related tasks such as table structure recognition, molecular structure recognition, music score recognition, as well as long fine-grained captioning and radiography report generation, on which PaliGemma 2 obtains state-of-the-art results. △ Less

Submitted 4 December, 2024; originally announced December 2024.

Wearable intelligent throat enables natural speech in stroke patients with dysarthria

Authors: Chenyu Tang, Shuo Gao, Cong Li, Wentian Yi, Yuxuan Jin, Xiaoxue Zhai, Sixuan Lei, Hongbei Meng, Zibo Zhang, Muzi Xu, Shengbo Wang, Xuhang Chen, Chenxi Wang, Hongyun Yang, Ningli Wang, Wenyu Wang, Jin Cao, Xiaodong Feng, Peter Smielewski, Yu Pan, Wenhui Song, Martin Birchall, Luigi G. Occhipinti

Abstract: Wearable silent speech systems hold significant potential for restoring communication in patients with speech impairments. However, seamless, coherent speech remains elusive, and clinical efficacy is still unproven. Here, we present an AI-driven intelligent throat (IT) system that integrates throat muscle vibrations and carotid pulse signal sensors with large language model (LLM) processing to ena… ▽ More Wearable silent speech systems hold significant potential for restoring communication in patients with speech impairments. However, seamless, coherent speech remains elusive, and clinical efficacy is still unproven. Here, we present an AI-driven intelligent throat (IT) system that integrates throat muscle vibrations and carotid pulse signal sensors with large language model (LLM) processing to enable fluent, emotionally expressive communication. The system utilizes ultrasensitive textile strain sensors to capture high-quality signals from the neck area and supports token-level processing for real-time, continuous speech decoding, enabling seamless, delay-free communication. In tests with five stroke patients with dysarthria, IT's LLM agents intelligently corrected token errors and enriched sentence-level emotional and logical coherence, achieving low error rates (4.2% word error rate, 2.9% sentence error rate) and a 55% increase in user satisfaction. This work establishes a portable, intuitive communication platform for patients with dysarthria with the potential to be applied broadly across different neurological conditions and in multi-language support systems. △ Less

Submitted 14 March, 2025; v1 submitted 27 November, 2024; originally announced November 2024.

Comments: 5 figures, 45 references

A Survey on LLM-as-a-Judge

Authors: Jiawei Gu, Xuhui Jiang, Zhichao Shi, Hexiang Tan, Xuehao Zhai, Chengjin Xu, Wei Li, Yinghan Shen, Shengjie Ma, Honghao Liu, Saizhuo Wang, Kun Zhang, Yuanzhuo Wang, Wen Gao, Lionel Ni, Jian Guo

Abstract: Accurate and consistent evaluation is crucial for decision-making across numerous fields, yet it remains a challenging task due to inherent subjectivity, variability, and scale. Large Language Models (LLMs) have achieved remarkable success across diverse domains, leading to the emergence of "LLM-as-a-Judge," where LLMs are employed as evaluators for complex tasks. With their ability to process div… ▽ More Accurate and consistent evaluation is crucial for decision-making across numerous fields, yet it remains a challenging task due to inherent subjectivity, variability, and scale. Large Language Models (LLMs) have achieved remarkable success across diverse domains, leading to the emergence of "LLM-as-a-Judge," where LLMs are employed as evaluators for complex tasks. With their ability to process diverse data types and provide scalable, cost-effective, and consistent assessments, LLMs present a compelling alternative to traditional expert-driven evaluations. However, ensuring the reliability of LLM-as-a-Judge systems remains a significant challenge that requires careful design and standardization. This paper provides a comprehensive survey of LLM-as-a-Judge, addressing the core question: How can reliable LLM-as-a-Judge systems be built? We explore strategies to enhance reliability, including improving consistency, mitigating biases, and adapting to diverse assessment scenarios. Additionally, we propose methodologies for evaluating the reliability of LLM-as-a-Judge systems, supported by a novel benchmark designed for this purpose. To advance the development and real-world deployment of LLM-as-a-Judge systems, we also discussed practical applications, challenges, and future directions. This survey serves as a foundational reference for researchers and practitioners in this rapidly evolving field. △ Less

Submitted 9 March, 2025; v1 submitted 23 November, 2024; originally announced November 2024.

Comments: Project Page: https://awesome-llm-as-a-judge.github.io/

Towards Next-Generation Medical Agent: How o1 is Reshaping Decision-Making in Medical Scenarios

Authors: Shaochen Xu, Yifan Zhou, Zhengliang Liu, Zihao Wu, Tianyang Zhong, Huaqin Zhao, Yiwei Li, Hanqi Jiang, Yi Pan, Junhao Chen, Jin Lu, Wei Zhang, Tuo Zhang, Lu Zhang, Dajiang Zhu, Xiang Li, Wei Liu, Quanzheng Li, Andrea Sikora, Xiaoming Zhai, Zhen Xiang, Tianming Liu

Abstract: Artificial Intelligence (AI) has become essential in modern healthcare, with large language models (LLMs) offering promising advances in clinical decision-making. Traditional model-based approaches, including those leveraging in-context demonstrations and those with specialized medical fine-tuning, have demonstrated strong performance in medical language processing but struggle with real-time adap… ▽ More Artificial Intelligence (AI) has become essential in modern healthcare, with large language models (LLMs) offering promising advances in clinical decision-making. Traditional model-based approaches, including those leveraging in-context demonstrations and those with specialized medical fine-tuning, have demonstrated strong performance in medical language processing but struggle with real-time adaptability, multi-step reasoning, and handling complex medical tasks. Agent-based AI systems address these limitations by incorporating reasoning traces, tool selection based on context, knowledge retrieval, and both short- and long-term memory. These additional features enable the medical AI agent to handle complex medical scenarios where decision-making should be built on real-time interaction with the environment. Therefore, unlike conventional model-based approaches that treat medical queries as isolated questions, medical AI agents approach them as complex tasks and behave more like human doctors. In this paper, we study the choice of the backbone LLM for medical AI agents, which is the foundation for the agent's overall reasoning and action generation. In particular, we consider the emergent o1 model and examine its impact on agents' reasoning, tool-use adaptability, and real-time information retrieval across diverse clinical scenarios, including high-stakes settings such as intensive care units (ICUs). Our findings demonstrate o1's ability to enhance diagnostic accuracy and consistency, paving the way for smarter, more responsive AI tools that support better patient outcomes and decision-making efficacy in clinical practice. △ Less

Submitted 16 November, 2024; originally announced November 2024.

Transcending Language Boundaries: Harnessing LLMs for Low-Resource Language Translation

Authors: Peng Shu, Junhao Chen, Zhengliang Liu, Hui Wang, Zihao Wu, Tianyang Zhong, Yiwei Li, Huaqin Zhao, Hanqi Jiang, Yi Pan, Yifan Zhou, Constance Owl, Xiaoming Zhai, Ninghao Liu, Claudio Saunt, Tianming Liu

Abstract: Large Language Models (LLMs) have demonstrated remarkable success across a wide range of tasks and domains. However, their performance in low-resource language translation, particularly when translating into these languages, remains underexplored. This gap poses significant challenges, as linguistic barriers hinder the cultural preservation and development of minority communities. To address this… ▽ More Large Language Models (LLMs) have demonstrated remarkable success across a wide range of tasks and domains. However, their performance in low-resource language translation, particularly when translating into these languages, remains underexplored. This gap poses significant challenges, as linguistic barriers hinder the cultural preservation and development of minority communities. To address this issue, this paper introduces a novel retrieval-based method that enhances translation quality for low-resource languages by focusing on key terms, which involves translating keywords and retrieving corresponding examples from existing data. To evaluate the effectiveness of this method, we conducted experiments translating from English into three low-resource languages: Cherokee, a critically endangered indigenous language of North America; Tibetan, a historically and culturally significant language in Asia; and Manchu, a language with few remaining speakers. Our comparison with the zero-shot performance of GPT-4o and LLaMA 3.1 405B, highlights the significant challenges these models face when translating into low-resource languages. In contrast, our retrieval-based method shows promise in improving both word-level accuracy and overall semantic understanding by leveraging existing resources more effectively. △ Less

Submitted 18 November, 2024; originally announced November 2024.

Using Generative AI and Multi-Agents to Provide Automatic Feedback

Authors: Shuchen Guo, Ehsan Latif, Yifan Zhou, Xuan Huang, Xiaoming Zhai

Abstract: This study investigates the use of generative AI and multi-agent systems to provide automatic feedback in educational contexts, particularly for student constructed responses in science assessments. The research addresses a key gap in the field by exploring how multi-agent systems, called AutoFeedback, can improve the quality of GenAI-generated feedback, overcoming known issues such as over-praise… ▽ More This study investigates the use of generative AI and multi-agent systems to provide automatic feedback in educational contexts, particularly for student constructed responses in science assessments. The research addresses a key gap in the field by exploring how multi-agent systems, called AutoFeedback, can improve the quality of GenAI-generated feedback, overcoming known issues such as over-praise and over-inference that are common in single-agent large language models (LLMs). The study developed a multi-agent system consisting of two AI agents: one for generating feedback and another for validating and refining it. The system was tested on a dataset of 240 student responses, and its performance was compared to that of a single-agent LLM. Results showed that AutoFeedback significantly reduced the occurrence of over-praise and over-inference errors, providing more accurate and pedagogically sound feedback. The findings suggest that multi-agent systems can offer a more reliable solution for generating automated feedback in educational settings, highlighting their potential for scalable and personalized learning support. These results have important implications for educators and researchers seeking to leverage AI in formative assessments, offering a pathway to more effective feedback mechanisms that enhance student learning outcomes. △ Less

Submitted 11 November, 2024; originally announced November 2024.

Large Language Models for Manufacturing

Authors: Yiwei Li, Huaqin Zhao, Hanqi Jiang, Yi Pan, Zhengliang Liu, Zihao Wu, Peng Shu, Jie Tian, Tianze Yang, Shaochen Xu, Yanjun Lyu, Parker Blenk, Jacob Pence, Jason Rupram, Eliza Banu, Ninghao Liu, Linbing Wang, Wenzhan Song, Xiaoming Zhai, Kenan Song, Dajiang Zhu, Beiwen Li, Xianqiao Wang, Tianming Liu

Abstract: The rapid advances in Large Language Models (LLMs) have the potential to transform manufacturing industry, offering new opportunities to optimize processes, improve efficiency, and drive innovation. This paper provides a comprehensive exploration of the integration of LLMs into the manufacturing domain, focusing on their potential to automate and enhance various aspects of manufacturing, from prod… ▽ More The rapid advances in Large Language Models (LLMs) have the potential to transform manufacturing industry, offering new opportunities to optimize processes, improve efficiency, and drive innovation. This paper provides a comprehensive exploration of the integration of LLMs into the manufacturing domain, focusing on their potential to automate and enhance various aspects of manufacturing, from product design and development to quality control, supply chain optimization, and talent management. Through extensive evaluations across multiple manufacturing tasks, we demonstrate the remarkable capabilities of state-of-the-art LLMs, such as GPT-4V, in understanding and executing complex instructions, extracting valuable insights from vast amounts of data, and facilitating knowledge sharing. We also delve into the transformative potential of LLMs in reshaping manufacturing education, automating coding processes, enhancing robot control systems, and enabling the creation of immersive, data-rich virtual environments through the industrial metaverse. By highlighting the practical applications and emerging use cases of LLMs in manufacturing, this paper aims to provide a valuable resource for professionals, researchers, and decision-makers seeking to harness the power of these technologies to address real-world challenges, drive operational excellence, and unlock sustainable growth in an increasingly competitive landscape. △ Less

Submitted 28 October, 2024; originally announced October 2024.

A Systematic Assessment of OpenAI o1-Preview for Higher Order Thinking in Education

Authors: Ehsan Latif, Yifan Zhou, Shuchen Guo, Yizhu Gao, Lehong Shi, Matthew Nayaaba, Gyeonggeon Lee, Liang Zhang, Arne Bewersdorff, Luyang Fang, Xiantong Yang, Huaqin Zhao, Hanqi Jiang, Haoran Lu, Jiaxi Li, Jichao Yu, Weihang You, Zhengliang Liu, Vincent Shung Liu, Hui Wang, Zihao Wu, Jin Lu, Fei Dou, Ping Ma, Ninghao Liu , et al. (2 additional authors not shown)

Abstract: As artificial intelligence (AI) continues to advance, it demonstrates capabilities comparable to human intelligence, with significant potential to transform education and workforce development. This study evaluates OpenAI o1-preview's ability to perform higher-order cognitive tasks across 14 dimensions, including critical thinking, systems thinking, computational thinking, design thinking, metacog… ▽ More As artificial intelligence (AI) continues to advance, it demonstrates capabilities comparable to human intelligence, with significant potential to transform education and workforce development. This study evaluates OpenAI o1-preview's ability to perform higher-order cognitive tasks across 14 dimensions, including critical thinking, systems thinking, computational thinking, design thinking, metacognition, data literacy, creative thinking, abstract reasoning, quantitative reasoning, logical reasoning, analogical reasoning, and scientific reasoning. We used validated instruments like the Ennis-Weir Critical Thinking Essay Test and the Biological Systems Thinking Test to compare the o1-preview's performance with human performance systematically. Our findings reveal that o1-preview outperforms humans in most categories, achieving 150% better results in systems thinking, computational thinking, data literacy, creative thinking, scientific reasoning, and abstract reasoning. However, compared to humans, it underperforms by around 25% in logical reasoning, critical thinking, and quantitative reasoning. In analogical reasoning, both o1-preview and humans achieved perfect scores. Despite these strengths, the o1-preview shows limitations in abstract reasoning, where human psychology students outperform it, highlighting the continued importance of human oversight in tasks requiring high-level abstraction. These results have significant educational implications, suggesting a shift toward developing human skills that complement AI, such as creativity, abstract reasoning, and critical thinking. This study emphasizes the transformative potential of AI in education and calls for a recalibration of educational goals, teaching methods, and curricula to align with an AI-driven world. △ Less

Submitted 11 October, 2024; originally announced October 2024.

Comments: An assessment of OpenAI o1-Preview for Higher Order Thinking in Education

Transformers are Efficient Compilers, Provably

Authors: Xiyu Zhai, Runlong Zhou, Liao Zhang, Simon Shaolei Du

Abstract: Transformer-based large language models (LLMs) have demonstrated surprisingly robust performance across a wide range of language-related tasks, including programming language understanding and generation. In this paper, we take the first steps towards a formal investigation of using transformers as compilers from an expressive power perspective. To this end, we introduce a representative programmi… ▽ More Transformer-based large language models (LLMs) have demonstrated surprisingly robust performance across a wide range of language-related tasks, including programming language understanding and generation. In this paper, we take the first steps towards a formal investigation of using transformers as compilers from an expressive power perspective. To this end, we introduce a representative programming language, Mini-Husky, which encapsulates key features of modern C-like languages. We show that if the input code sequence has a bounded depth in both the Abstract Syntax Tree (AST) and type inference (reasonable assumptions based on the clean code principle), then the number of parameters required by transformers depends only on the logarithm of the input sequence length to handle compilation tasks, such as AST construction, symbol resolution, and type analysis. A significant technical challenge stems from the fact that transformers operate at a low level, where each layer processes the input sequence as raw vectors without explicitly associating them with predefined structure or meaning. In contrast, high-level compiler tasks necessitate managing intricate relationships and structured program information. Our primary technical contribution is the development of a domain-specific language, Cybertron, which generates formal proofs of the transformer's expressive power, scaling to address compiler tasks. We further establish that recurrent neural networks (RNNs) require at least a linear number of parameters relative to the input sequence, leading to an exponential separation between transformers and RNNs. Finally, we empirically validate our theoretical results by comparing transformers and RNNs on compiler tasks within Mini-Husky. △ Less

Submitted 24 January, 2025; v1 submitted 7 October, 2024; originally announced October 2024.

Comments: 65 pages

MSC Class: 68T07 (Primary)

Transforming Teachers' Roles and Agencies in the Era of Generative AI: Perceptions, Acceptance, Knowledge, and Practices

Authors: Xiaoming Zhai

Abstract: This paper explores the transformative impact of Generative Artificial Intelligence (GenAI) on teachers' roles and agencies in education, presenting a comprehensive framework that addresses teachers' perceptions, knowledge, acceptance, and practices of GenAI. As GenAI technologies, such as ChatGPT, become increasingly integrated into educational settings, teachers are required to adapt to evolving… ▽ More This paper explores the transformative impact of Generative Artificial Intelligence (GenAI) on teachers' roles and agencies in education, presenting a comprehensive framework that addresses teachers' perceptions, knowledge, acceptance, and practices of GenAI. As GenAI technologies, such as ChatGPT, become increasingly integrated into educational settings, teachers are required to adapt to evolving classroom dynamics, where AI plays a significant role in content creation, personalized learning, and student engagement. However, existing literature often treats these factors in isolation, overlooking how they collectively influence teachers' ability to effectively integrate GenAI into their pedagogical practices. This paper fills this gap by proposing a framework that categorizes teachers into four roles -- Observer, Adopter, Collaborator, and Innovator -- each representing different levels of GenAI engagement, outlining teachers' agencies in GenAI classrooms. By highlighting the need for continuous professional development and institutional support, we demonstrate how teachers can evolve from basic GenAI users to co-creators of knowledge alongside GenAI systems. The findings emphasize that for GenAI to reach its full educational potential, teachers must not only accept and understand its capabilities but also integrate it deeply into their teaching strategies. This study contributes to the growing literature on GenAI in education, offering practical implications for supporting teachers in navigating the complexities of GenAI adoption. △ Less

Submitted 3 October, 2024; originally announced October 2024.

Lost-in-Distance: Impact of Contextual Proximity on LLM Performance in Graph Tasks

Authors: Hamed Firooz, Maziar Sanjabi, Wenlong Jiang, Xiaoling Zhai

Abstract: Despite significant advancements, Large Language Models (LLMs) exhibit blind spots that impair their ability to retrieve and process relevant contextual data effectively. We demonstrate that LLM performance in graph tasks with complexities beyond the "needle-in-a-haystack" scenario-where solving the problem requires cross-referencing and reasoning across multiple subproblems jointly-is influenced… ▽ More Despite significant advancements, Large Language Models (LLMs) exhibit blind spots that impair their ability to retrieve and process relevant contextual data effectively. We demonstrate that LLM performance in graph tasks with complexities beyond the "needle-in-a-haystack" scenario-where solving the problem requires cross-referencing and reasoning across multiple subproblems jointly-is influenced by the proximity of relevant information within the context, a phenomenon we term "lost-in-distance". We examine two fundamental graph tasks: identifying common connections between two nodes and assessing similarity among three nodes, and show that the model's performance in these tasks significantly depends on the relative positioning of common edges. We evaluate three publicly available LLMs using various graph encoding techniques that represent graph structures for LLM input. We propose a formulation for the lost-in-distance phenomenon and demonstrate that lost-in-distance and lost-in-the middle phenomenas occur independently. Results indicate that model accuracy can decline by up to 6x as the distance between node connections increases, independent of graph encoding and model size. △ Less

Submitted 1 January, 2025; v1 submitted 2 October, 2024; originally announced October 2024.

Evaluation of OpenAI o1: Opportunities and Challenges of AGI

Authors: Tianyang Zhong, Zhengliang Liu, Yi Pan, Yutong Zhang, Yifan Zhou, Shizhe Liang, Zihao Wu, Yanjun Lyu, Peng Shu, Xiaowei Yu, Chao Cao, Hanqi Jiang, Hanxu Chen, Yiwei Li, Junhao Chen, Huawen Hu, Yiheng Liu, Huaqin Zhao, Shaochen Xu, Haixing Dai, Lin Zhao, Ruidong Zhang, Wei Zhao, Zhenyuan Yang, Jingyuan Chen , et al. (50 additional authors not shown)

Abstract: This comprehensive study evaluates the performance of OpenAI's o1-preview large language model across a diverse array of complex reasoning tasks, spanning multiple domains, including computer science, mathematics, natural sciences, medicine, linguistics, and social sciences. Through rigorous testing, o1-preview demonstrated remarkable capabilities, often achieving human-level or superior performan… ▽ More This comprehensive study evaluates the performance of OpenAI's o1-preview large language model across a diverse array of complex reasoning tasks, spanning multiple domains, including computer science, mathematics, natural sciences, medicine, linguistics, and social sciences. Through rigorous testing, o1-preview demonstrated remarkable capabilities, often achieving human-level or superior performance in areas ranging from coding challenges to scientific reasoning and from language processing to creative problem-solving. Key findings include: -83.3% success rate in solving complex competitive programming problems, surpassing many human experts. -Superior ability in generating coherent and accurate radiology reports, outperforming other evaluated models. -100% accuracy in high school-level mathematical reasoning tasks, providing detailed step-by-step solutions. -Advanced natural language inference capabilities across general and specialized domains like medicine. -Impressive performance in chip design tasks, outperforming specialized models in areas such as EDA script generation and bug analysis. -Remarkable proficiency in anthropology and geology, demonstrating deep understanding and reasoning in these specialized fields. -Strong capabilities in quantitative investing. O1 has comprehensive financial knowledge and statistical modeling skills. -Effective performance in social media analysis, including sentiment analysis and emotion recognition. The model excelled particularly in tasks requiring intricate reasoning and knowledge integration across various fields. While some limitations were observed, including occasional errors on simpler problems and challenges with certain highly specialized concepts, the overall results indicate significant progress towards artificial general intelligence. △ Less

Submitted 7 July, 2025; v1 submitted 27 September, 2024; originally announced September 2024.

OpenRANet: Neuralized Spectrum Access by Joint Subcarrier and Power Allocation with Optimization-based Deep Learning

Authors: Siya Chen, Chee Wei Tan, Xiangping Zhai, H. Vincent Poor

Abstract: The next-generation radio access network (RAN), known as Open RAN, is poised to feature an AI-native interface for wireless cellular networks, including emerging satellite-terrestrial systems, making deep learning integral to its operation. In this paper, we address the nonconvex optimization challenge of joint subcarrier and power allocation in Open RAN, with the objective of minimizing the total… ▽ More The next-generation radio access network (RAN), known as Open RAN, is poised to feature an AI-native interface for wireless cellular networks, including emerging satellite-terrestrial systems, making deep learning integral to its operation. In this paper, we address the nonconvex optimization challenge of joint subcarrier and power allocation in Open RAN, with the objective of minimizing the total power consumption while ensuring users meet their transmission data rate requirements. We propose OpenRANet, an optimization-based deep learning model that integrates machine-learning techniques with iterative optimization algorithms. We start by transforming the original nonconvex problem into convex subproblems through decoupling, variable transformation, and relaxation techniques. These subproblems are then efficiently solved using iterative methods within the standard interference function framework, enabling the derivation of primal-dual solutions. These solutions integrate seamlessly as a convex optimization layer within OpenRANet, enhancing constraint adherence, solution accuracy, and computational efficiency by combining machine learning with convex analysis, as shown in numerical experiments. OpenRANet also serves as a foundation for designing resource-constrained AI-native wireless optimization strategies for broader scenarios like multi-cell systems, satellite-terrestrial networks, and future Open RAN deployments with complex power consumption requirements. △ Less

Submitted 10 February, 2025; v1 submitted 31 August, 2024; originally announced September 2024.

Comments: This paper has been accepted by the IEEE Transactions on Green Communications and Networking

Cap2Sum: Learning to Summarize Videos by Generating Captions

Authors: Cairong Zhao, Chutian Wang, Zifan Song, Guosheng Hu, Haonan Chen, Xiaofan Zhai

Abstract: With the rapid growth of video data on the internet, video summarization is becoming a very important AI technology. However, due to the high labelling cost of video summarization, existing studies have to be conducted on small-scale datasets, leading to limited performance and generalization capacity. In this work, we introduce the use of dense video captions as a supervision signal to train vide… ▽ More With the rapid growth of video data on the internet, video summarization is becoming a very important AI technology. However, due to the high labelling cost of video summarization, existing studies have to be conducted on small-scale datasets, leading to limited performance and generalization capacity. In this work, we introduce the use of dense video captions as a supervision signal to train video summarization models. Motivated by this, we propose Cap2Sum, a model that learns to summarize videos by generating captions, to exploit dense video caption annotations. This weakly-supervised approach allows us to train the models on large-scale dense video caption datasets to achieve better performance and generalization capacity. To further improve the generalization capacity, we introduce a CLIP (a strong vision-language model) Prior mechanism to enhance the learning of important objects that captions may ignore in the videos. In practice, Cap2Sum can perform zero-shot video summarization or be fine-tuned by the ground-truth summary or video caption of the target dataset. To examine the performance of Cap2Sum after weakly-supervised fine-tuning by the video captions, we propose two new datasets, TVSum-Caption and SumMe-Caption, which are derived from two common video summarization datasets and will be publicly released. We conduct extensive experiments and the results demonstrate that our method achieves significant improvements in performance and generalization capacity compared with previous methods. △ Less

Submitted 22 August, 2024; originally announced August 2024.

Comments: 13 pages, 4 figures

Modeling Electromagnetic Signal Injection Attacks on Camera-based Smart Systems: Applications and Mitigation

Authors: Youqian Zhang, Michael Cheung, Chunxi Yang, Xinwei Zhai, Zitong Shen, Xinyu Ji, Eugene Y. Fu, Sze-Yiu Chau, Xiapu Luo

Abstract: Numerous safety- or security-critical systems depend on cameras to perceive their surroundings, further allowing artificial intelligence (AI) to analyze the captured images to make important decisions. However, a concerning attack vector has emerged, namely, electromagnetic waves, which pose a threat to the integrity of these systems. Such attacks enable attackers to manipulate the images remotely… ▽ More Numerous safety- or security-critical systems depend on cameras to perceive their surroundings, further allowing artificial intelligence (AI) to analyze the captured images to make important decisions. However, a concerning attack vector has emerged, namely, electromagnetic waves, which pose a threat to the integrity of these systems. Such attacks enable attackers to manipulate the images remotely, leading to incorrect AI decisions, e.g., autonomous vehicles missing detecting obstacles ahead resulting in collisions. The lack of understanding regarding how different systems react to such attacks poses a significant security risk. Furthermore, no effective solutions have been demonstrated to mitigate this threat. To address these gaps, we modeled the attacks and developed a simulation method for generating adversarial images. Through rigorous analysis, we confirmed that the effects of the simulated adversarial images are indistinguishable from those from real attacks. This method enables researchers and engineers to rapidly assess the susceptibility of various AI vision applications to these attacks, without the need for constructing complicated attack devices. In our experiments, most of the models demonstrated vulnerabilities to these attacks, emphasizing the need to enhance their robustness. Fortunately, our modeling and simulation method serves as a stepping stone toward developing more resilient models. We present a pilot study on adversarial training to improve their robustness against attacks, and our results demonstrate a significant improvement by recovering up to 91% performance, offering a promising direction for mitigating this threat. △ Less

Submitted 9 August, 2024; originally announced August 2024.

Comments: 13 pages, 10 figures, 4 tables

Unveiling Scoring Processes: Dissecting the Differences between LLMs and Human Graders in Automatic Scoring

Authors: Xuansheng Wu, Padmaja Pravin Saraf, Gyeonggeon Lee, Ehsan Latif, Ninghao Liu, Xiaoming Zhai

Abstract: Large language models (LLMs) have demonstrated strong potential in performing automatic scoring for constructed response assessments. While constructed responses graded by humans are usually based on given grading rubrics, the methods by which LLMs assign scores remain largely unclear. It is also uncertain how closely AI's scoring process mirrors that of humans or if it adheres to the same grading… ▽ More Large language models (LLMs) have demonstrated strong potential in performing automatic scoring for constructed response assessments. While constructed responses graded by humans are usually based on given grading rubrics, the methods by which LLMs assign scores remain largely unclear. It is also uncertain how closely AI's scoring process mirrors that of humans or if it adheres to the same grading criteria. To address this gap, this paper uncovers the grading rubrics that LLMs used to score students' written responses to science tasks and their alignment with human scores. We also examine whether enhancing the alignments can improve scoring accuracy. Specifically, we prompt LLMs to generate analytic rubrics that they use to assign scores and study the alignment gap with human grading rubrics. Based on a series of experiments with various configurations of LLM settings, we reveal a notable alignment gap between human and LLM graders. While LLMs can adapt quickly to scoring tasks, they often resort to shortcuts, bypassing deeper logical reasoning expected in human grading. We found that incorporating high-quality analytical rubrics designed to reflect human grading logic can mitigate this gap and enhance LLMs' scoring accuracy. These results underscore the need for a nuanced approach when applying LLMs in science education and highlight the importance of aligning LLM outputs with human expectations to ensure efficient and accurate automatic scoring. △ Less

Submitted 21 February, 2025; v1 submitted 4 July, 2024; originally announced July 2024.

Comments: Accepted by Technology, Knowledge, and Learning (TKNL)

Generative AI as a Learning Buddy and Teaching Assistant: Pre-service Teachers' Uses and Attitudes

Authors: Matthew Nyaaba, Lehong Shi, Macharious Nabang, Xiaoming Zhai, Patrick Kyeremeh, Samuel Arthur Ayoberd, Bismark Nyaaba Akanzire

Abstract: To uncover pre-service teachers' (PSTs') user experience and perceptions of generative artificial intelligence (GenAI) applications, we surveyed 167 Ghana PSTs' specific uses of GenAI as a learning buddy and teaching assistant, and their attitudes towards these applications. Employing exploratory factor analysis (EFA), we identified three key factors shaping PSTs' attitudes towards GenAI: teaching… ▽ More To uncover pre-service teachers' (PSTs') user experience and perceptions of generative artificial intelligence (GenAI) applications, we surveyed 167 Ghana PSTs' specific uses of GenAI as a learning buddy and teaching assistant, and their attitudes towards these applications. Employing exploratory factor analysis (EFA), we identified three key factors shaping PSTs' attitudes towards GenAI: teaching, learning, and ethical and advocacy factors. The mean scores of these factors revealed a generally positive attitude towards GenAI, indicating high levels of agreement on its potential to enhance PSTs' content knowledge and access to learning and teaching resources, thereby reducing their need for assistance from colleagues. Specifically, PSTs use GenAI as a learning buddy to access reading materials, in-depth content explanations, and practical examples, and as a teaching assistant to enhance teaching resources, develop assessment strategies, and plan lessons. A regression analysis showed that background factors such as age, gender, and year of study do not predict PSTs' attitudes towards GenAI, but age and year of study significantly predict the frequency of their use of GenAI, while gender does not. These findings suggest that older PSTs and those further along in their teacher education programs may use GenAI more frequently, but their perceptions of the application remain unchanged. However, PSTs expressed concerns about the accuracy and trustworthiness of the information provided by GenAI applications. We, therefore, suggest addressing these concerns to ensure PSTs can confidently rely on these applications in their teacher preparation programs. Additionally, we recommend targeted strategies to integrate GenAI more effectively into both learning and teaching processes for PSTs. △ Less

Submitted 3 June, 2024; originally announced July 2024.

PaliGemma: A versatile 3B VLM for transfer

Authors: Lucas Beyer, Andreas Steiner, André Susano Pinto, Alexander Kolesnikov, Xiao Wang, Daniel Salz, Maxim Neumann, Ibrahim Alabdulmohsin, Michael Tschannen, Emanuele Bugliarello, Thomas Unterthiner, Daniel Keysers, Skanda Koppula, Fangyu Liu, Adam Grycner, Alexey Gritsenko, Neil Houlsby, Manoj Kumar, Keran Rong, Julian Eisenschlos, Rishabh Kabra, Matthias Bauer, Matko Bošnjak, Xi Chen, Matthias Minderer , et al. (10 additional authors not shown)

Abstract: PaliGemma is an open Vision-Language Model (VLM) that is based on the SigLIP-So400m vision encoder and the Gemma-2B language model. It is trained to be a versatile and broadly knowledgeable base model that is effective to transfer. It achieves strong performance on a wide variety of open-world tasks. We evaluate PaliGemma on almost 40 diverse tasks including standard VLM benchmarks, but also more… ▽ More PaliGemma is an open Vision-Language Model (VLM) that is based on the SigLIP-So400m vision encoder and the Gemma-2B language model. It is trained to be a versatile and broadly knowledgeable base model that is effective to transfer. It achieves strong performance on a wide variety of open-world tasks. We evaluate PaliGemma on almost 40 diverse tasks including standard VLM benchmarks, but also more specialized tasks such as remote-sensing and segmentation. △ Less

Submitted 10 October, 2024; v1 submitted 10 July, 2024; originally announced July 2024.

Comments: v2 adds Appendix H and I and a few citations

Image-Conditional Diffusion Transformer for Underwater Image Enhancement

Authors: Xingyang Nie, Su Pan, Xiaoyu Zhai, Shifei Tao, Fengzhong Qu, Biao Wang, Huilin Ge, Guojie Xiao

Abstract: Underwater image enhancement (UIE) has attracted much attention owing to its importance for underwater operation and marine engineering. Motivated by the recent advance in generative models, we propose a novel UIE method based on image-conditional diffusion transformer (ICDT). Our method takes the degraded underwater image as the conditional input and converts it into latent space where ICDT is ap… ▽ More Underwater image enhancement (UIE) has attracted much attention owing to its importance for underwater operation and marine engineering. Motivated by the recent advance in generative models, we propose a novel UIE method based on image-conditional diffusion transformer (ICDT). Our method takes the degraded underwater image as the conditional input and converts it into latent space where ICDT is applied. ICDT replaces the conventional U-Net backbone in a denoising diffusion probabilistic model (DDPM) with a transformer, and thus inherits favorable properties such as scalability from transformers. Furthermore, we train ICDT with a hybrid loss function involving variances to achieve better log-likelihoods, which meanwhile significantly accelerates the sampling process. We experimentally assess the scalability of ICDTs and compare with prior works in UIE on the Underwater ImageNet dataset. Besides good scaling properties, our largest model, ICDT-XL/2, outperforms all comparison methods, achieving state-of-the-art (SOTA) quality of image enhancement. △ Less

Submitted 7 July, 2024; originally announced July 2024.

Fast, Scalable, Energy-Efficient Non-element-wise Matrix Multiplication on FPGA

Authors: Xuqi Zhu, Huaizhi Zhang, JunKyu Lee, Jiacheng Zhu, Chandrajit Pal, Sangeet Saha, Klaus D. McDonald-Maier, Xiaojun Zhai

Abstract: Modern Neural Network (NN) architectures heavily rely on vast numbers of multiply-accumulate arithmetic operations, constituting the predominant computational cost. Therefore, this paper proposes a high-throughput, scalable and energy efficient non-element-wise matrix multiplication unit on FPGAs as a basic component of the NNs. We firstly streamline inter-layer and intra-layer redundancies of MAD… ▽ More Modern Neural Network (NN) architectures heavily rely on vast numbers of multiply-accumulate arithmetic operations, constituting the predominant computational cost. Therefore, this paper proposes a high-throughput, scalable and energy efficient non-element-wise matrix multiplication unit on FPGAs as a basic component of the NNs. We firstly streamline inter-layer and intra-layer redundancies of MADDNESS algorithm, a LUT-based approximate matrix multiplication, to design a fast, efficient scalable approximate matrix multiplication module termed "Approximate Multiplication Unit (AMU)". The AMU optimizes LUT-based matrix multiplications further through dedicated memory management and access design, decoupling computational overhead from input resolution and boosting FPGA-based NN accelerator efficiency significantly. The experimental results show that using our AMU achieves up to 9x higher throughput and 112x higher energy efficiency over the state-of-the-art solutions for the FPGA-based Quantised Neural Network (QNN) accelerators. △ Less

Submitted 7 July, 2024; v1 submitted 2 July, 2024; originally announced July 2024.

Toward a Diffusion-Based Generalist for Dense Vision Tasks

Authors: Yue Fan, Yongqin Xian, Xiaohua Zhai, Alexander Kolesnikov, Muhammad Ferjad Naeem, Bernt Schiele, Federico Tombari

Abstract: Building generalized models that can solve many computer vision tasks simultaneously is an intriguing direction. Recent works have shown image itself can be used as a natural interface for general-purpose visual perception and demonstrated inspiring results. In this paper, we explore diffusion-based vision generalists, where we unify different types of dense prediction tasks as conditional image g… ▽ More Building generalized models that can solve many computer vision tasks simultaneously is an intriguing direction. Recent works have shown image itself can be used as a natural interface for general-purpose visual perception and demonstrated inspiring results. In this paper, we explore diffusion-based vision generalists, where we unify different types of dense prediction tasks as conditional image generation and re-purpose pre-trained diffusion models for it. However, directly applying off-the-shelf latent diffusion models leads to a quantization issue. Thus, we propose to perform diffusion in pixel space and provide a recipe for finetuning pre-trained text-to-image diffusion models for dense vision tasks. In experiments, we evaluate our method on four different types of tasks and show competitive performance to the other vision generalists. △ Less

Submitted 29 June, 2024; originally announced July 2024.

Comments: Published at CVPR 2024 as a workshop paper