Datasets:

AdithyaSK

/

Arxiv_AI_Research_Papers

like 0

2405.19255

Haowen Xu

Jose Tupayachi, Haowen Xu, Olufemi A. Omitaomu, Mustafa Can Camur, Aliza Sharmin, Xueping Li

Towards Next-Generation Urban Decision Support Systems through AI-Powered Generation of Scientific Ontology using Large Language Models -- A Case in Optimizing Intermodal Freight Transportation

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

The incorporation of Artificial Intelligence (AI) models into various optimization systems is on the rise. Yet, addressing complex urban and environmental management problems normally requires in-depth domain science and informatics expertise. This expertise is essential for deriving data and simulation-driven for informed decision support. In this context, we investigate the potential of leveraging the pre-trained Large Language Models (LLMs). By adopting ChatGPT API as the reasoning core, we outline an integrated workflow that encompasses natural language processing, methontology-based prompt tuning, and transformers. This workflow automates the creation of scenario-based ontology using existing research articles and technical manuals of urban datasets and simulations. The outcomes of our methodology are knowledge graphs in widely adopted ontology languages (e.g., OWL, RDF, SPARQL). These facilitate the development of urban decision support systems by enhancing the data and metadata modeling, the integration of complex datasets, the coupling of multi-domain simulation models, and the formulation of decision-making metrics and workflow. The feasibility of our methodology is evaluated through a comparative analysis that juxtaposes our AI-generated ontology with the well-known Pizza Ontology employed in tutorials for popular ontology software (e.g., prot\'eg\'e). We close with a real-world case study of optimizing the complex urban system of multi-modal freight transportation by generating anthologies of various domain data and simulations to support informed decision-making.

2405.19444

Zhenwen Liang

Zhenwen Liang, Dian Yu, Wenhao Yu, Wenlin Yao, Zhihan Zhang, Xiangliang Zhang, Dong Yu

MathChat: Benchmarking Mathematical Reasoning and Instruction Following in Multi-Turn Interactions

cs.AI

http://creativecommons.org/licenses/by/4.0/

Large language models (LLMs) have demonstrated impressive capabilities in mathematical problem solving, particularly in single turn question answering formats. However, real world scenarios often involve mathematical question answering that requires multi turn or interactive information exchanges, and the performance of LLMs on these tasks is still underexplored. This paper introduces MathChat, a comprehensive benchmark specifically designed to evaluate LLMs across a broader spectrum of mathematical tasks. These tasks are structured to assess the models' abilities in multiturn interactions and open ended generation. We evaluate the performance of various SOTA LLMs on the MathChat benchmark, and we observe that while these models excel in single turn question answering, they significantly underperform in more complex scenarios that require sustained reasoning and dialogue understanding. To address the above limitations of existing LLMs when faced with multiturn and open ended tasks, we develop MathChat sync, a synthetic dialogue based math dataset for LLM finetuning, focusing on improving models' interaction and instruction following capabilities in conversations. Experimental results emphasize the need for training LLMs with diverse, conversational instruction tuning datasets like MathChatsync. We believe this work outlines one promising direction for improving the multiturn mathematical reasoning abilities of LLMs, thus pushing forward the development of LLMs that are more adept at interactive mathematical problem solving and real world applications.

2405.19453

Chamani Shiranthika Jayakody Kankanamalage

Chamani Shiranthika, Parvaneh Saeedi, Ivan V. Baji\'c

Optimizing Split Points for Error-Resilient SplitFed Learning

Accepted for poster presentation at the Women in Computer Vision (WiCV) workshop in CVPR 2024

cs.AI

http://creativecommons.org/publicdomain/zero/1.0/

Recent advancements in decentralized learning, such as Federated Learning (FL), Split Learning (SL), and Split Federated Learning (SplitFed), have expanded the potentials of machine learning. SplitFed aims to minimize the computational burden on individual clients in FL and parallelize SL while maintaining privacy. This study investigates the resilience of SplitFed to packet loss at model split points. It explores various parameter aggregation strategies of SplitFed by examining the impact of splitting the model at different points-either shallow split or deep split-on the final global model performance. The experiments, conducted on a human embryo image segmentation task, reveal a statistically significant advantage of a deeper split point.

2405.19456

Xisen Wang

Xisen Wang, Yigit Ihlamur

An Automated Startup Evaluation Pipeline: Startup Success Forecasting Framework (SSFF)

For relevant code: https://github.com/Xisen-Wang/Startup-Success-Forecasting-Framework

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Evaluating startups in their early stages is a complex task that requires detailed analysis by experts. While automating this process on a large scale can significantly impact businesses, the inherent complexity poses challenges. This paper addresses this challenge by introducing the Startup Success Forecasting Framework (SSFF), a new automated system that combines traditional machine learning with advanced language models. This intelligent agent-based architecture is designed to reason, act, synthesize, and decide like a venture capitalist to perform the analysis end-to-end. The SSFF is made up of three main parts: - Prediction Block: Uses random forests and neural networks to make predictions. - Analyst Block: Simulates VC analysis scenario and uses SOTA prompting techniques - External Knowledge Block: Gathers real-time information from external sources. This framework requires minimal input data about the founder and startup description, enhances it with additional data from external resources, and performs a detailed analysis with high accuracy, all in an automated manner

2405.19464

Haowen Xu

Haowen Xu, Femi Omitaomu, Soheil Sabri, Xiao Li, Yongze Song

Leveraging Generative AI for Smart City Digital Twins: A Survey on the Autonomous Generation of Data, Scenarios, 3D City Models, and Urban Designs

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

The digital transformation of modern cities by integrating advanced information, communication, and computing technologies has marked the epoch of data-driven smart city applications for efficient and sustainable urban management. Despite their effectiveness, these applications often rely on massive amounts of high-dimensional and multi-domain data for monitoring and characterizing different urban sub-systems, presenting challenges in application areas that are limited by data quality and availability, as well as costly efforts for generating urban scenarios and design alternatives. As an emerging research area in deep learning, Generative Artificial Intelligence (AI) models have demonstrated their unique values in data and code generation. This survey paper aims to explore the innovative integration of generative AI techniques and urban digital twins to address challenges in the realm of smart cities in various urban sectors, such as transportation and mobility management, energy system operations, building and infrastructure management, and urban design. The survey starts with the introduction of popular generative AI models with their application areas, followed by a structured review of the existing urban science applications that leverage the autonomous capability of the generative AI techniques to facilitate (a) data augmentation for promoting urban monitoring and predictive analytics, (b) synthetic data and scenario generation, (c) automated 3D city modeling, and (d) generative urban design and optimization. Based on the review, this survey discusses potential opportunities and technical strategies that integrate generative AI models into the next-generation urban digital twins for more reliable, scalable, and automated management of smart cities.

2405.19498

Robert Johansson

Robert Johansson

Machine Psychology: Integrating Operant Conditioning with the Non-Axiomatic Reasoning System for Advancing Artificial General Intelligence Research

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

This paper introduces an interdisciplinary framework called Machine Psychology, which merges principles from operant learning psychology with a specific Artificial Intelligence model, the Non-Axiomatic Reasoning System (NARS), to enhance Artificial General Intelligence (AGI) research. The core premise of this framework is that adaptation is crucial to both biological and artificial intelligence and can be understood through operant conditioning principles. The study assesses this approach via three operant learning tasks using OpenNARS for Applications (ONA): simple discrimination, changing contingencies, and conditional discrimination tasks. In the simple discrimination task, NARS demonstrated rapid learning, achieving perfect accuracy during both training and testing phases. The changing contingencies task showcased NARS's adaptability, as it successfully adjusted its behavior when task conditions were reversed. In the conditional discrimination task, NARS handled complex learning scenarios effectively, achieving high accuracy by forming and utilizing intricate hypotheses based on conditional cues. These findings support the application of operant conditioning as a framework for creating adaptive AGI systems. NARS's ability to operate under conditions of insufficient knowledge and resources, coupled with its sensorimotor reasoning capabilities, establishes it as a robust model for AGI. The Machine Psychology framework, by incorporating elements of natural intelligence such as continuous learning and goal-driven behavior, offers a scalable and flexible approach for real-world applications. Future research should investigate using enhanced NARS systems, more advanced tasks, and applying this framework to diverse, complex challenges to further progress the development of human-level AI.

2405.19522

Nestor Maslej

Nestor Maslej, Loredana Fattorini, Raymond Perrault, Vanessa Parli, Anka Reuel, Erik Brynjolfsson, John Etchemendy, Katrina Ligett, Terah Lyons, James Manyika, Juan Carlos Niebles, Yoav Shoham, Russell Wald, and Jack Clark

Artificial Intelligence Index Report 2024

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

The 2024 Index is our most comprehensive to date and arrives at an important moment when AI's influence on society has never been more pronounced. This year, we have broadened our scope to more extensively cover essential trends such as technical advancements in AI, public perceptions of the technology, and the geopolitical dynamics surrounding its development. Featuring more original data than ever before, this edition introduces new estimates on AI training costs, detailed analyses of the responsible AI landscape, and an entirely new chapter dedicated to AI's impact on science and medicine. The AI Index report tracks, collates, distills, and visualizes data related to artificial intelligence (AI). Our mission is to provide unbiased, rigorously vetted, broadly sourced data in order for policymakers, researchers, executives, journalists, and the general public to develop a more thorough and nuanced understanding of the complex field of AI. The AI Index is recognized globally as one of the most credible and authoritative sources for data and insights on artificial intelligence. Previous editions have been cited in major newspapers, including the The New York Times, Bloomberg, and The Guardian, have amassed hundreds of academic citations, and been referenced by high-level policymakers in the United States, the United Kingdom, and the European Union, among other places. This year's edition surpasses all previous ones in size, scale, and scope, reflecting the growing significance that AI is coming to hold in all of our lives.

2405.19606

Zhuang Qi

Xiaming Che, Junlin Zhang, Zhuang Qi, Xin Qi

Relation Modeling and Distillation for Learning with Noisy Labels

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Learning with noisy labels has become an effective strategy for enhancing the robustness of models, which enables models to better tolerate inaccurate data. Existing methods either focus on optimizing the loss function to mitigate the interference from noise, or design procedures to detect potential noise and correct errors. However, their effectiveness is often compromised in representation learning due to the dilemma where models overfit to noisy labels. To address this issue, this paper proposes a relation modeling and distillation framework that models inter-sample relationships via self-supervised learning and employs knowledge distillation to enhance understanding of latent associations, which mitigate the impact of noisy labels. Specifically, the proposed method, termed RMDNet, includes two main modules, where the relation modeling (RM) module implements the contrastive learning technique to learn representations of all data, an unsupervised approach that effectively eliminates the interference of noisy tags on feature extraction. The relation-guided representation learning (RGRL) module utilizes inter-sample relation learned from the RM module to calibrate the representation distribution for noisy samples, which is capable of improving the generalization of the model in the inference phase. Notably, the proposed RMDNet is a plug-and-play framework that can integrate multiple methods to its advantage. Extensive experiments were conducted on two datasets, including performance comparison, ablation study, in-depth analysis and case study. The results show that RMDNet can learn discriminative representations for noisy data, which results in superior performance than the existing methods.

2405.19631

Akul Goel

Akul Goel, Surya Narayanan Hari, Belinda Waltman, Matt Thomson

Leveraging Open-Source Large Language Models for encoding Social Determinants of Health using an Intelligent Router

cs.AI

http://creativecommons.org/licenses/by/4.0/

Social Determinants of Health (SDOH) play a significant role in patient health outcomes. The Center of Disease Control (CDC) introduced a subset of ICD-10 codes called Z-codes in an attempt to officially recognize and measure SDOH in the health care system. However, these codes are rarely annotated in a patient's Electronic Health Record (EHR), and instead, in many cases, need to be inferred from clinical notes. Previous research has shown that large language models (LLMs) show promise on extracting unstructured data from EHRs. However, with thousands of models to choose from with unique architectures and training sets, it's difficult to choose one model that performs the best on coding tasks. Further, clinical notes contain trusted health information making the use of closed-source language models from commercial vendors difficult, so the identification of open source LLMs that can be run within health organizations and exhibits high performance on SDOH tasks is an urgent problem. Here, we introduce an intelligent routing system for SDOH coding that uses a language model router to direct medical record data to open source LLMs that demonstrate optimal performance on specific SDOH codes. The intelligent routing system exhibits state of the art performance of 97.4% accuracy averaged across 5 codes, including homelessness and food insecurity, on par with closed models such as GPT-4o. In order to train the routing system and validate models, we also introduce a synthetic data generation and validation paradigm to increase the scale of training data without needing privacy protected medical records. Together, we demonstrate an architecture for intelligent routing of inputs to task-optimal language models to achieve high performance across a set of medical coding sub-tasks.

2405.19642

Mengjie Gan

Mengjie Gan, Penglong Lian, Zhiheng Su, Jiyang Zhang, Jialong Huang, Benhao Wang, Jianxiao Zou and Shicai Fan

Few-shot fault diagnosis based on multi-scale graph convolution filtering for industry

6 pages, 2 figures, 2 tables, 63rd IEEE Conference on Decision and Control

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Industrial equipment fault diagnosis often encounter challenges such as the scarcity of fault data, complex operating conditions, and varied types of failures. Signal analysis, data statistical learning, and conventional deep learning techniques face constraints under these conditions due to their substantial data requirements and the necessity for transfer learning to accommodate new failure modes. To effectively leverage information and extract the intrinsic characteristics of faults across different domains under limited sample conditions, this paper introduces a fault diagnosis approach employing Multi-Scale Graph Convolution Filtering (MSGCF). MSGCF enhances the traditional Graph Neural Network (GNN) framework by integrating both local and global information fusion modules within the graph convolution filter block. This advancement effectively mitigates the over-smoothing issue associated with excessive layering of graph convolutional layers while preserving a broad receptive field. It also reduces the risk of overfitting in few-shot diagnosis, thereby augmenting the model's representational capacity. Experiments on the University of Paderborn bearing dataset (PU) demonstrate that the MSGCF method proposed herein surpasses alternative approaches in accuracy, thereby offering valuable insights for industrial fault diagnosis in few-shot learning scenarios.

2405.19654

Jinxia Yang

Jinxia Yang, Bing Su, Wayne Xin Zhao, Ji-Rong Wen

Unlocking the Power of Spatial and Temporal Information in Medical Multimodal Pre-training

Accepted at ICML 2024

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Medical vision-language pre-training methods mainly leverage the correspondence between paired medical images and radiological reports. Although multi-view spatial images and temporal sequences of image-report pairs are available in off-the-shelf multi-modal medical datasets, most existing methods have not thoroughly tapped into such extensive supervision signals. In this paper, we introduce the Med-ST framework for fine-grained spatial and temporal modeling to exploit information from multiple spatial views of chest radiographs and temporal historical records. For spatial modeling, Med-ST employs the Mixture of View Expert (MoVE) architecture to integrate different visual features from both frontal and lateral views. To achieve a more comprehensive alignment, Med-ST not only establishes the global alignment between whole images and texts but also introduces modality-weighted local alignment between text tokens and spatial regions of images. For temporal modeling, we propose a novel cross-modal bidirectional cycle consistency objective by forward mapping classification (FMC) and reverse mapping regression (RMR). By perceiving temporal information from simple to complex, Med-ST can learn temporal semantics. Experimental results across four distinct tasks demonstrate the effectiveness of Med-ST, especially in temporal classification tasks. Our code and model are available at https://github.com/SVT-Yang/MedST.

2405.19656

Han Liu

Han Liu, Peng Cui, Bingning Wang, Jun Zhu, Xiaolin Hu

Accurate and Reliable Predictions with Mutual-Transport Ensemble

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Deep Neural Networks (DNNs) have achieved remarkable success in a variety of tasks, especially when it comes to prediction accuracy. However, in complex real-world scenarios, particularly in safety-critical applications, high accuracy alone is not enough. Reliable uncertainty estimates are crucial. Modern DNNs, often trained with cross-entropy loss, tend to be overconfident, especially with ambiguous samples. To improve uncertainty calibration, many techniques have been developed, but they often compromise prediction accuracy. To tackle this challenge, we propose the ``mutual-transport ensemble'' (MTE). This approach introduces a co-trained auxiliary model and adaptively regularizes the cross-entropy loss using Kullback-Leibler (KL) divergence between the prediction distributions of the primary and auxiliary models. We conducted extensive studies on various benchmarks to validate the effectiveness of our method. The results show that MTE can simultaneously enhance both accuracy and uncertainty calibration. For example, on the CIFAR-100 dataset, our MTE method on ResNet34/50 achieved significant improvements compared to previous state-of-the-art method, with absolute accuracy increases of 2.4%/3.7%, relative reductions in ECE of $42.3%/29.4%, and relative reductions in classwise-ECE of 11.6%/15.3%.

2405.19686

Jingwei Sun

Jingwei Sun, Zhixu Du, Yiran Chen

Knowledge Graph Tuning: Real-time Large Language Model Personalization based on Human Feedback

cs.AI

http://creativecommons.org/licenses/by/4.0/

Large language models (LLMs) have demonstrated remarkable proficiency in a range of natural language processing tasks. Once deployed, LLMs encounter users with personalized factual knowledge, and such personalized knowledge is consistently reflected through users' interactions with the LLMs. To enhance user experience, real-time model personalization is essential, allowing LLMs to adapt user-specific knowledge based on user feedback during human-LLM interactions. Existing methods mostly require back-propagation to finetune the model parameters, which incurs high computational and memory costs. In addition, these methods suffer from low interpretability, which will cause unforeseen impacts on model performance during long-term use, where the user's personalized knowledge is accumulated extensively.To address these challenges, we propose Knowledge Graph Tuning (KGT), a novel approach that leverages knowledge graphs (KGs) to personalize LLMs. KGT extracts personalized factual knowledge triples from users' queries and feedback and optimizes KGs without modifying the LLM parameters. Our method improves computational and memory efficiency by avoiding back-propagation and ensures interpretability by making the KG adjustments comprehensible to humans.Experiments with state-of-the-art LLMs, including GPT-2, Llama2, and Llama3, show that KGT significantly improves personalization performance while reducing latency and GPU memory costs. Ultimately, KGT offers a promising solution of effective, efficient, and interpretable real-time LLM personalization during user interactions with the LLMs.

2405.19694

Wenjing Xie

Wenjing Xie, Juxin Niu, Chun Jason Xue, Nan Guan

Grade Like a Human: Rethinking Automated Assessment with Large Language Models

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

While large language models (LLMs) have been used for automated grading, they have not yet achieved the same level of performance as humans, especially when it comes to grading complex questions. Existing research on this topic focuses on a particular step in the grading procedure: grading using predefined rubrics. However, grading is a multifaceted procedure that encompasses other crucial steps, such as grading rubrics design and post-grading review. There has been a lack of systematic research exploring the potential of LLMs to enhance the entire grading~process. In this paper, we propose an LLM-based grading system that addresses the entire grading procedure, including the following key components: 1) Developing grading rubrics that not only consider the questions but also the student answers, which can more accurately reflect students' performance. 2) Under the guidance of grading rubrics, providing accurate and consistent scores for each student, along with customized feedback. 3) Conducting post-grading review to better ensure accuracy and fairness. Additionally, we collected a new dataset named OS from a university operating system course and conducted extensive experiments on both our new dataset and the widely used Mohler dataset. Experiments demonstrate the effectiveness of our proposed approach, providing some new insights for developing automated grading systems based on LLMs.

2405.19736

Yunlong Liu

Dayang Liang, Jinyang Lai, and Yunlong Liu

Learning Task-relevant Sequence Representations via Intrinsic Dynamics Characteristics in Reinforcement Learning

cs.AI

http://creativecommons.org/licenses/by/4.0/

Learning task-relevant state representations is crucial to solving the problem of scene generalization in visual deep reinforcement learning. Prior work typically establishes a self-supervised auxiliary learner, introducing elements (e.g., rewards and actions) to extract task-relevant state information from observations through behavioral similarity metrics. However, the methods often ignore the inherent relationships between the elements (e.g., dynamics relationships) that are essential for learning accurate representations, and they are also limited to single-step metrics, which impedes the discrimination of short-term similar task/behavior information in long-term dynamics transitions. To solve the issues, we propose an intrinsic dynamic characteristics-driven sequence representation learning method (DSR) over a common DRL frame. Concretely, inspired by the fact of state transition in the underlying system, it constrains the optimization of the encoder via modeling the dynamics equations related to the state transition, which prompts the latent encoding information to satisfy the state transition process and thereby distinguishes state space and noise space. Further, to refine the ability of encoding similar tasks based on dynamics constraints, DSR also sequentially models inherent dynamics equation relationships from the perspective of sequence elements' frequency domain and multi-step prediction. Finally, experimental results show that DSR has achieved a significant performance boost in the Distracting DMControl Benchmark, with an average of 78.9% over the backbone baseline. Further results indicate that it also achieves the best performance in real-world autonomous driving tasks in the CARLA simulator. Moreover, the qualitative analysis results of t-SNE visualization validate that our method possesses superior representation ability on visual tasks.

2405.19761

Zhihao Chang

Zhihao Chang, Linzhu Yu, Huan Li, Sai Wu, Gang Chen, Dongxiang Zhang

Revisiting CNNs for Trajectory Similarity Learning

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Similarity search is a fundamental but expensive operator in querying trajectory data, due to its quadratic complexity of distance computation. To mitigate the computational burden for long trajectories, neural networks have been widely employed for similarity learning and each trajectory is encoded as a high-dimensional vector for similarity search with linear complexity. Given the sequential nature of trajectory data, previous efforts have been primarily devoted to the utilization of RNNs or Transformers. In this paper, we argue that the common practice of treating trajectory as sequential data results in excessive attention to capturing long-term global dependency between two sequences. Instead, our investigation reveals the pivotal role of local similarity, prompting a revisit of simple CNNs for trajectory similarity learning. We introduce ConvTraj, incorporating both 1D and 2D convolutions to capture sequential and geo-distribution features of trajectories, respectively. In addition, we conduct a series of theoretical analyses to justify the effectiveness of ConvTraj. Experimental results on three real-world large-scale datasets demonstrate that ConvTraj achieves state-of-the-art accuracy in trajectory similarity search. Owing to the simple network structure of ConvTraj, the training and inference speed on the Porto dataset with 1.6 million trajectories are increased by at least $240$x and $2.16$x, respectively. The source code and dataset can be found at \textit{\url{https://github.com/Proudc/ConvTraj}}.

2405.19808

Herman Cappelen

Herman Cappelen and Josh Dever

AI with Alien Content and Alien Metasemantics

20 pages, book chapter

in Ernie Lepore and Luvell Anderson (Eds), The Oxford Handbook of Applied Philosophy of Language, Oxford Handbooks (2024)

10.1093/oxfordhb/9780192844118.013.47

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

AlphaGo plays chess and Go in a creative and novel way. It is natural for us to attribute contents to it, such as that it doesn't view being several pawns behind, if it has more board space, as bad. The framework introduced in Cappelen and Dever (2021) provides a way of thinking about the semantics and the metasemantics of AI content: does AlphaGo entertain contents like this, and if so, in virtue of what does a given state of the program mean that particular content? One salient question Cappelen and Dever didn't consider was the possibility of alien content. Alien content is content that is not or cannot be expressed by human beings. It's highly plausible that AlphaGo, or any other sophisticated AI system, expresses alien contents. That this is so, moreover, is plausibly a metasemantic fact: a fact that has to do with how AI comes to entertain content in the first place, one that will heed the vastly different etiology of AI and human content. This chapter explores the question of alien content in AI from a semantic and metasemantic perspective. It lays out the logical space of possible responses to the semantic and metasemantic questions alien content poses, considers whether and how we humans could communicate with entities who express alien content, and points out that getting clear about such questions might be important for more 'applied' issues in the philosophy of AI, such as existential risk and XAI.

2405.19816

Sylvain Chevallier

Manon Verbockhaven (TAU, LISN), Sylvain Chevallier (TAU, LISN), Guillaume Charpiat (TAU, LISN)

Growing Tiny Networks: Spotting Expressivity Bottlenecks and Fixing Them Optimally

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Machine learning tasks are generally formulated as optimization problems, where one searches for an optimal function within a certain functional space. In practice, parameterized functional spaces are considered, in order to be able to perform gradient descent. Typically, a neural network architecture is chosen and fixed, and its parameters (connection weights) are optimized, yielding an architecture-dependent result. This way of proceeding however forces the evolution of the function during training to lie within the realm of what is expressible with the chosen architecture, and prevents any optimization across architectures. Costly architectural hyper-parameter optimization is often performed to compensate for this. Instead, we propose to adapt the architecture on the fly during training. We show that the information about desirable architectural changes, due to expressivity bottlenecks when attempting to follow the functional gradient, can be extracted from %the backpropagation. To do this, we propose a mathematical definition of expressivity bottlenecks, which enables us to detect, quantify and solve them while training, by adding suitable neurons when and where needed. Thus, while the standard approach requires large networks, in terms of number of neurons per layer, for expressivity and optimization reasons, we are able to start with very small neural networks and let them grow appropriately. As a proof of concept, we show results~on the CIFAR dataset, matching large neural network accuracy, with competitive training time, while removing the need for standard architectural hyper-parameter search.

2405.19832

Herman Cappelen

Herman Cappelen, Josh Dever and John Hawthorne

AI Safety: A Climb To Armageddon?

20 page article

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper presents an argument that certain AI safety measures, rather than mitigating existential risk, may instead exacerbate it. Under certain key assumptions - the inevitability of AI failure, the expected correlation between an AI system's power at the point of failure and the severity of the resulting harm, and the tendency of safety measures to enable AI systems to become more powerful before failing - safety efforts have negative expected utility. The paper examines three response strategies: Optimism, Mitigation, and Holism. Each faces challenges stemming from intrinsic features of the AI safety landscape that we term Bottlenecking, the Perfection Barrier, and Equilibrium Fluctuation. The surprising robustness of the argument forces a re-examination of core assumptions around AI safety and points to several avenues for further research.

2405.19837

Margarida Romero

Margarida Romero (LINE, COMUE UCA, ULaval, Mnemosyne)

Lifelong learning challenges in the era of artificial intelligence: a computational thinking perspective

IRMBAM, Ipag, Jul 2024, Nice, France

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The rapid advancement of artificial intelligence (AI) has brought significant challenges to the education and workforce skills required to take advantage of AI for human-AI collaboration in the workplace. As AI continues to reshape industries and job markets, the need to define how AI literacy can be considered in lifelong learning has become increasingly critical (Cetindamar et al., 2022; Laupichler et al., 2022; Romero et al., 2023). Like any new technology, AI is the subject of both hopes and fears, and what it entails today presents major challenges (Cugurullo \& Acheampong, 2023; Villani et al., 2018). It also raises profound questions about our own humanity. Will the machine surpass the intelligence of the humans who designed it? What will be the relationship between so-called AI and our human intelligences? How could human-AI collaboration be regulated in a way that serves the Sustainable Development Goals (SDGs)? This paper provides a review of the challenges of lifelong learning in the era of AI from a computational thinking, critical thinking, and creative competencies perspective, highlighting the implications for management and leadership in organizations.

2405.19850

Yuxiao Luo

Yuxiao Luo, Zhongcai Cao, Xin Jin, Kang Liu, Ling Yin

Deciphering Human Mobility: Inferring Semantics of Trajectories with Large Language Models

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Understanding human mobility patterns is essential for various applications, from urban planning to public safety. The individual trajectory such as mobile phone location data, while rich in spatio-temporal information, often lacks semantic detail, limiting its utility for in-depth mobility analysis. Existing methods can infer basic routine activity sequences from this data, lacking depth in understanding complex human behaviors and users' characteristics. Additionally, they struggle with the dependency on hard-to-obtain auxiliary datasets like travel surveys. To address these limitations, this paper defines trajectory semantic inference through three key dimensions: user occupation category, activity sequence, and trajectory description, and proposes the Trajectory Semantic Inference with Large Language Models (TSI-LLM) framework to leverage LLMs infer trajectory semantics comprehensively and deeply. We adopt spatio-temporal attributes enhanced data formatting (STFormat) and design a context-inclusive prompt, enabling LLMs to more effectively interpret and infer the semantics of trajectory data. Experimental validation on real-world trajectory datasets demonstrates the efficacy of TSI-LLM in deciphering complex human mobility patterns. This study explores the potential of LLMs in enhancing the semantic analysis of trajectory data, paving the way for more sophisticated and accessible human mobility research.

2405.19915

Huihong Shi

Huihong Shi, Xin Cheng, Wendong Mao, and Zhongfeng Wang

P$^2$-ViT: Power-of-Two Post-Training Quantization and Acceleration for Fully Quantized Vision Transformer

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Vision Transformers (ViTs) have excelled in computer vision tasks but are memory-consuming and computation-intensive, challenging their deployment on resource-constrained devices. To tackle this limitation, prior works have explored ViT-tailored quantization algorithms but retained floating-point scaling factors, which yield non-negligible re-quantization overhead, limiting ViTs' hardware efficiency and motivating more hardware-friendly solutions. To this end, we propose \emph{P$^2$-ViT}, the first \underline{P}ower-of-Two (PoT) \underline{p}ost-training quantization and acceleration framework to accelerate fully quantized ViTs. Specifically, {as for quantization,} we explore a dedicated quantization scheme to effectively quantize ViTs with PoT scaling factors, thus minimizing the re-quantization overhead. Furthermore, we propose coarse-to-fine automatic mixed-precision quantization to enable better accuracy-efficiency trade-offs. {In terms of hardware,} we develop {a dedicated chunk-based accelerator} featuring multiple tailored sub-processors to individually handle ViTs' different types of operations, alleviating reconfigurable overhead. Additionally, we design {a tailored row-stationary dataflow} to seize the pipeline processing opportunity introduced by our PoT scaling factors, thereby enhancing throughput. Extensive experiments consistently validate P$^2$-ViT's effectiveness. {Particularly, we offer comparable or even superior quantization performance with PoT scaling factors when compared to the counterpart with floating-point scaling factors. Besides, we achieve up to $\mathbf{10.1\times}$ speedup and $\mathbf{36.8\times}$ energy saving over GPU's Turing Tensor Cores, and up to $\mathbf{1.84\times}$ higher computation utilization efficiency against SOTA quantization-based ViT accelerators. Codes are available at \url{https://github.com/shihuihong214/P2-ViT}.

2405.19946

Xuanfa Jin

Xuanfa Jin, Ziyan Wang, Yali Du, Meng Fang, Haifeng Zhang, Jun Wang

Learning to Discuss Strategically: A Case Study on One Night Ultimate Werewolf

27 pages, 5 figures

cs.AI

http://creativecommons.org/licenses/by/4.0/

Communication is a fundamental aspect of human society, facilitating the exchange of information and beliefs among people. Despite the advancements in large language models (LLMs), recent agents built with these often neglect the control over discussion tactics, which are essential in communication scenarios and games. As a variant of the famous communication game Werewolf, One Night Ultimate Werewolf (ONUW) requires players to develop strategic discussion policies due to the potential role changes that increase the uncertainty and complexity of the game. In this work, we first present the existence of the Perfect Bayesian Equilibria (PBEs) in two scenarios of the ONUW game: one with discussion and one without. The results showcase that the discussion greatly changes players' utilities by affecting their beliefs, emphasizing the significance of discussion tactics. Based on the insights obtained from the analyses, we propose an RL-instructed language agent framework, where a discussion policy trained by reinforcement learning (RL) is employed to determine appropriate discussion tactics to adopt. Our experimental results on several ONUW game settings demonstrate the effectiveness and generalizability of our proposed framework.

2405.19956

Jing Wen

Jing Wen

HOLMES: to Detect Adversarial Examples with Multiple Detectors

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

Deep neural networks (DNNs) can easily be cheated by some imperceptible but purposeful noise added to images, and erroneously classify them. Previous defensive work mostly focused on retraining the models or detecting the noise, but has either shown limited success rates or been attacked by new adversarial examples. Instead of focusing on adversarial images or the interior of DNN models, we observed that adversarial examples generated by different algorithms can be identified based on the output of DNNs (logits). Logit can serve as an exterior feature to train detectors. Then, we propose HOLMES (Hierarchically Organized Light-weight Multiple dEtector System) to reinforce DNNs by detecting potential adversarial examples to minimize the threats they may bring in practical. HOLMES is able to distinguish \textit{unseen} adversarial examples from multiple attacks with high accuracy and low false positive rates than single detector systems even in an adaptive model. To ensure the diversity and randomness of detectors in HOLMES, we use two methods: training dedicated detectors for each label and training detectors with top-k logits. Our effective and inexpensive strategies neither modify original DNN models nor require its internal parameters. HOLMES is not only compatible with all kinds of learning models (even only with external APIs), but also complementary to other defenses to achieve higher detection rates (may also fully protect the system against various adversarial examples).

2405.19970

Petra Bayerl

Petra Saskia Bayerl, Babak Akhgar, Ernesto La Mattina, Barbara Pirillo, Ioana Cotoi, Davide Ariu, Matteo Mauri, Jorge Garcia, Dimitris Kavallieros, Antonia Kardara, Konstantina Karagiorgou

Strategies to Counter Artificial Intelligence in Law Enforcement: Cross-Country Comparison of Citizens in Greece, Italy and Spain

20th International Conference on Information and Knowledge Engineering (IKE'21), 3 papges, 1 figure

cs.AI

http://creativecommons.org/licenses/by/4.0/

This paper investigates citizens' counter-strategies to the use of Artificial Intelligence (AI) by law enforcement agencies (LEAs). Based on information from three countries (Greece, Italy and Spain) we demonstrate disparities in the likelihood of ten specific counter-strategies. We further identified factors that increase the propensity for counter-strategies. Our study provides an important new perspective to societal impacts of security-focused AI applications by illustrating the conscious, strategic choices by citizens when confronted with AI capabilities for LEAs.

2405.20046

Zhuang Qi

Zhuang Qi, Lei Meng, Weihao He, Ruohan Zhang, Yu Wang, Xin Qi, Xiangxu Meng

Cross-Training with Multi-View Knowledge Fusion for Heterogenous Federated Learning

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Federated learning benefits from cross-training strategies, which enables models to train on data from distinct sources to improve the generalization capability. However, the data heterogeneity between sources may lead models to gradually forget previously acquired knowledge when undergoing cross-training to adapt to new tasks or data sources. We argue that integrating personalized and global knowledge to gather information from multiple perspectives could potentially improve performance. To achieve this goal, this paper presents a novel approach that enhances federated learning through a cross-training scheme incorporating multi-view information. Specifically, the proposed method, termed FedCT, includes three main modules, where the consistency-aware knowledge broadcasting module aims to optimize model assignment strategies, which enhances collaborative advantages between clients and achieves an efficient federated learning process. The multi-view knowledge-guided representation learning module leverages fused prototypical knowledge from both global and local views to enhance the preservation of local knowledge before and after model exchange, as well as to ensure consistency between local and global knowledge. The mixup-based feature augmentation module aggregates rich information to further increase the diversity of feature spaces, which enables the model to better discriminate complex samples. Extensive experiments were conducted on four datasets in terms of performance comparison, ablation study, in-depth analysis and case study. The results demonstrated that FedCT alleviates knowledge forgetting from both local and global views, which enables it outperform state-of-the-art methods.

2405.20121

Dong Caiyin

Sun Zhanbo, Dong Caiyin, Ji Ang, Zhao Ruibin, Zhao Yu

A Structure-Aware Lane Graph Transformer Model for Vehicle Trajectory Prediction

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Accurate prediction of future trajectories for surrounding vehicles is vital for the safe operation of autonomous vehicles. This study proposes a Lane Graph Transformer (LGT) model with structure-aware capabilities. Its key contribution lies in encoding the map topology structure into the attention mechanism. To address variations in lane information from different directions, four Relative Positional Encoding (RPE) matrices are introduced to capture the local details of the map topology structure. Additionally, two Shortest Path Distance (SPD) matrices are employed to capture distance information between two accessible lanes. Numerical results indicate that the proposed LGT model achieves a significantly higher prediction performance on the Argoverse 2 dataset. Specifically, the minFDE$_6$ metric was decreased by 60.73% compared to the Argoverse 2 baseline model (Nearest Neighbor) and the b-minFDE$_6$ metric was reduced by 2.65% compared to the baseline LaneGCN model. Furthermore, ablation experiments demonstrated that the consideration of map topology structure led to a 4.24% drop in the b-minFDE$_6$ metric, validating the effectiveness of this model.

2405.20138

Toshio Suzuki

Fuki Ito, Toshio Suzuki

Separation and Collapse of Equilibria Inequalities on AND-OR Trees without Shape Constraints

42 pages, 1 figure

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Herein, we investigate the randomized complexity, which is the least cost against the worst input, of AND-OR tree computation by imposing various restrictions on the algorithm to find the Boolean value of the root of that tree and no restrictions on the tree shape. When a tree satisfies a certain condition regarding its symmetry, directional algorithms proposed by Saks and Wigderson (1986), special randomized algorithms, are known to achieve the randomized complexity. Furthermore, there is a known example of a tree that is so unbalanced that no directional algorithm achieves the randomized complexity (Vereshchagin 1998). In this study, we aim to identify where deviations arise between the general randomized Boolean decision tree and its special case, directional algorithms. In this paper, we show that for any AND-OR tree, randomized depth-first algorithms, which form a broader class compared with directional algorithms, have the same equilibrium as that of the directional algorithms. Thus, we get the collapse result on equilibria inequalities that holds for an arbitrary AND-OR tree. This implies that there exists a case where even depth-first algorithms cannot be the fastest, leading to the separation result on equilibria inequality. Additionally, a new algorithm is introduced as a key concept for proof of the separation result.

2405.20142

Jingjing Guo

Chao Zhang, Weirong Cui, and Jingjing Guo

MSSC-BiMamba: Multimodal Sleep Stage Classification and Early Diagnosis of Sleep Disorders with Bidirectional Mamba

10 pages

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Monitoring sleep states is essential for evaluating sleep quality and diagnosing sleep disorders. Traditional manual staging is time-consuming and prone to subjective bias, often resulting in inconsistent outcomes. Here, we developed an automated model for sleep staging and disorder classification to enhance diagnostic accuracy and efficiency. Considering the characteristics of polysomnography (PSG) multi-lead sleep monitoring, we designed a multimodal sleep state classification model, MSSC-BiMamba, that combines an Efficient Channel Attention (ECA) mechanism with a Bidirectional State Space Model (BSSM). The ECA module allows for weighting data from different sensor channels, thereby amplifying the influence of diverse sensor inputs. Additionally, the implementation of bidirectional Mamba (BiMamba) enables the model to effectively capture the multidimensional features and long-range dependencies of PSG data. The developed model demonstrated impressive performance on sleep stage classification tasks on both the ISRUC-S3 and ISRUC-S1 datasets, respectively containing data with healthy and unhealthy sleep patterns. Also, the model exhibited a high accuracy for sleep health prediction when evaluated on a combined dataset consisting of ISRUC and Sleep-EDF. Our model, which can effectively handle diverse sleep conditions, is the first to apply BiMamba to sleep staging with multimodal PSG data, showing substantial gains in computational and memory efficiency over traditional Transformer-style models. This method enhances sleep health management by making monitoring more accessible and extending advanced healthcare through innovative technology.

2405.20202

Ke Yi

Ke Yi, Yuhui Xu, Heng Chang, Chen Tang, Yuan Meng, Tong Zhang, Jia Li

One QuantLLM for ALL: Fine-tuning Quantized LLMs Once for Efficient Deployments

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Large Language Models (LLMs) have advanced rapidly but face significant memory demands. While quantization has shown promise for LLMs, current methods typically require lengthy training to alleviate the performance degradation from quantization loss. However, deploying LLMs across diverse scenarios with different resource constraints, e.g., servers and personal computers, requires repeated training per application, which amplifies the lengthy training problem. Given that, it is advantageous to train a once-for-all (OFA) supernet capable of yielding diverse optimal subnets for downstream applications through one-shot training. Nonetheless, the scale of current language models impedes efficiency and amplifies interference from weight sharing between subnets. We make an initial attempt to extend the once-for-all framework to large language models. Specifically, we decouple shared weights to eliminate the interference and incorporate Low-Rank adapters for training efficiency. Furthermore, we observe the imbalance allocation of training resources from the traditional uniform sampling. A non-parametric scheduler is introduced to adjust the sampling rate for each quantization configuration, achieving a more balanced allocation among subnets with varying demands. We validate the approach on LLaMA2 families, and downstream evaluation confirms our ability to maintain high performance while significantly reducing deployment time faced with multiple scenarios.

2405.20234

Cheng'an Wei

Cheng'an Wei, Kai Chen, Yue Zhao, Yujia Gong, Lu Xiang, and Shenchen Zhu

Context Injection Attacks on Large Language Models

cs.AI

http://creativecommons.org/licenses/by/4.0/

Large Language Models (LLMs) such as ChatGPT and Llama-2 have become prevalent in real-world applications, exhibiting impressive text generation performance. LLMs are fundamentally developed from a scenario where the input data remains static and lacks a clear structure. To behave interactively over time, LLM-based chat systems must integrate additional contextual information (i.e., chat history) into their inputs, following a pre-defined structure. This paper identifies how such integration can expose LLMs to misleading context from untrusted sources and fail to differentiate between system and user inputs, allowing users to inject context. We present a systematic methodology for conducting context injection attacks aimed at eliciting disallowed responses by introducing fabricated context. This could lead to illegal actions, inappropriate content, or technology misuse. Our context fabrication strategies, acceptance elicitation and word anonymization, effectively create misleading contexts that can be structured with attacker-customized prompt templates, achieving injection through malicious user messages. Comprehensive evaluations on real-world LLMs such as ChatGPT and Llama-2 confirm the efficacy of the proposed attack with success rates reaching 97%. We also discuss potential countermeasures that can be adopted for attack detection and developing more secure models. Our findings provide insights into the challenges associated with the real-world deployment of LLMs for interactive and structured data scenarios.

2405.20421

Qianqi Yan

Qianqi Yan, Xuehai He, Xiang Yue, Xin Eric Wang

Worse than Random? An Embarrassingly Simple Probing Evaluation of Large Multimodal Models in Medical VQA

cs.AI

http://creativecommons.org/licenses/by/4.0/

Large Multimodal Models (LMMs) have shown remarkable progress in the field of medical Visual Question Answering (Med-VQA), achieving high accuracy on existing benchmarks. However, their reliability under robust evaluation is questionable. This study reveals that state-of-the-art models, when subjected to simple probing evaluation, perform worse than random guessing on medical diagnosis questions. To address this critical evaluation problem, we introduce the Probing Evaluation for Medical Diagnosis (ProbMed) dataset to rigorously assess LMM performance in medical imaging through probing evaluation and procedural diagnosis. Particularly, probing evaluation features pairing original questions with negation questions with hallucinated attributes, while procedural diagnosis requires reasoning across various diagnostic dimensions for each image, including modality recognition, organ identification, clinical findings, abnormalities, and positional grounding. Our evaluation reveals that top-performing models like GPT-4V and Gemini Pro perform worse than random guessing on specialized diagnostic questions, indicating significant limitations in handling fine-grained medical inquiries. Besides, models like LLaVA-Med struggle even with more general questions, and results from CheXagent demonstrate the transferability of expertise across different modalities of the same organ, showing that specialized domain knowledge is still crucial for improving performance. This study underscores the urgent need for more robust evaluation to ensure the reliability of LMMs in critical fields like medical diagnosis, and current LMMs are still far from applicable to those fields.

2405.20487

Yuta Kawakami

Yuta Kawakami, Manabu Kuroki, Jin Tian

Probabilities of Causation for Continuous and Vector Variables

cs.AI

http://creativecommons.org/licenses/by/4.0/

Probabilities of causation (PoC) are valuable concepts for explainable artificial intelligence and practical decision-making. PoC are originally defined for scalar binary variables. In this paper, we extend the concept of PoC to continuous treatment and outcome variables, and further generalize PoC to capture causal effects between multiple treatments and multiple outcomes. In addition, we consider PoC for a sub-population and PoC with multi-hypothetical terms to capture more sophisticated counterfactual information useful for decision-making. We provide a nonparametric identification theorem for each type of PoC we introduce. Finally, we illustrate the application of our results on a real-world dataset about education.

2405.20519

Shreyas Kapur

Shreyas Kapur, Erik Jenner, Stuart Russell

Diffusion On Syntax Trees For Program Synthesis

https://tree-diffusion.github.io

cs.AI

http://creativecommons.org/licenses/by-sa/4.0/

Large language models generate code one token at a time. Their autoregressive generation process lacks the feedback of observing the program's output. Training LLMs to suggest edits directly can be challenging due to the scarcity of rich edit data. To address these problems, we propose neural diffusion models that operate on syntax trees of any context-free grammar. Similar to image diffusion models, our method also inverts ``noise'' applied to syntax trees. Rather than generating code sequentially, we iteratively edit it while preserving syntactic validity, which makes it easy to combine this neural model with search. We apply our approach to inverse graphics tasks, where our model learns to convert images into programs that produce those images. Combined with search, our model is able to write graphics programs, see the execution result, and debug them to meet the required specifications. We additionally show how our system can write graphics programs for hand-drawn sketches.

2405.20600

Huiguang He

Kaicheng Fu, Changde Du, Xiaoyu Chen, Jie Peng, Huiguang He

Multi-label Class Incremental Emotion Decoding with Augmented Emotional Semantics Learning

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

Emotion decoding plays an important role in affective human-computer interaction. However, previous studies ignored the dynamic real-world scenario, where human experience a blend of multiple emotions which are incrementally integrated into the model, leading to the multi-label class incremental learning (MLCIL) problem. Existing methods have difficulty in solving MLCIL issue due to notorious catastrophic forgetting caused by partial label problem and inadequate label semantics mining. In this paper, we propose an augmented emotional semantics learning framework for multi-label class incremental emotion decoding. Specifically, we design an augmented emotional relation graph module with label disambiguation to handle the past-missing partial label problem. Then, we leverage domain knowledge from affective dimension space to alleviate future-missing partial label problem by knowledge distillation. Besides, an emotional semantics learning module is constructed with a graph autoencoder to obtain emotion embeddings in order to guide the semantic-specific feature decoupling for better multi-label learning. Extensive experiments on three datasets show the superiority of our method for improving emotion decoding performance and mitigating forgetting on MLCIL problem.

2405.20625

Mudit Verma

Atharva Gundawar, Mudit Verma, Lin Guan, Karthik Valmeekam, Siddhant Bhambri, Subbarao Kambhampati

Robust Planning with LLM-Modulo Framework: Case Study in Travel Planning

cs.AI

http://creativecommons.org/licenses/by/4.0/

As the applicability of Large Language Models (LLMs) extends beyond traditional text processing tasks, there is a burgeoning interest in their potential to excel in planning and reasoning assignments, realms traditionally reserved for System 2 cognitive competencies. Despite their perceived versatility, the research community is still unraveling effective strategies to harness these models in such complex domains. The recent discourse introduced by the paper on LLM Modulo marks a significant stride, proposing a conceptual framework that enhances the integration of LLMs into diverse planning and reasoning activities. This workshop paper delves into the practical application of this framework within the domain of travel planning, presenting a specific instance of its implementation. We are using the Travel Planning benchmark by the OSU NLP group, a benchmark for evaluating the performance of LLMs in producing valid itineraries based on user queries presented in natural language. While popular methods of enhancing the reasoning abilities of LLMs such as Chain of Thought, ReAct, and Reflexion achieve a meager 0%, 0.6%, and 0% with GPT3.5-Turbo respectively, our operationalization of the LLM-Modulo framework for TravelPlanning domain provides a remarkable improvement, enhancing baseline performances by 4.6x for GPT4-Turbo and even more for older models like GPT3.5-Turbo from 0% to 5%. Furthermore, we highlight the other useful roles of LLMs in the planning pipeline, as suggested in LLM-Modulo, which can be reliably operationalized such as extraction of useful critics and reformulator for critics.

2405.20653

Jiahao Yu

Jiahao Yu, Haozheng Luo, Jerry Yao-Chieh Hu, Wenbo Guo, Han Liu, Xinyu Xing

Enhancing Jailbreak Attack Against Large Language Models through Silent Tokens

cs.AI

http://creativecommons.org/licenses/by/4.0/

Along with the remarkable successes of Language language models, recent research also started to explore the security threats of LLMs, including jailbreaking attacks. Attackers carefully craft jailbreaking prompts such that a target LLM will respond to the harmful question. Existing jailbreaking attacks require either human experts or leveraging complicated algorithms to craft jailbreaking prompts. In this paper, we introduce BOOST, a simple attack that leverages only the eos tokens. We demonstrate that rather than constructing complicated jailbreaking prompts, the attacker can simply append a few eos tokens to the end of a harmful question. It will bypass the safety alignment of LLMs and lead to successful jailbreaking attacks. We further apply BOOST to four representative jailbreak methods and show that the attack success rates of these methods can be significantly enhanced by simply adding eos tokens to the prompt. To understand this simple but novel phenomenon, we conduct empirical analyses. Our analysis reveals that adding eos tokens makes the target LLM believe the input is much less harmful, and eos tokens have low attention values and do not affect LLM's understanding of the harmful questions, leading the model to actually respond to the questions. Our findings uncover how fragile an LLM is against jailbreak attacks, motivating the development of strong safety alignment approaches.

2405.20656

Javier Naranjo-Alcazar

Javier Naranjo-Alcazar, Jordi Grau-Haro, Pedro Zuccarello, David Almenar, Jesus Lopez-Ballester

Automatic Counting and Classification of Mosquito Eggs in Field Traps

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The analysis of the field traps where the mosquitoes insert their eggs is vital to check that the sterile insect technique (SIT) is working properly. This is because the number of hatched eggs may indicate that the sterile males are not competing with the wild ones. Nowadays, the study of the traps is done manually by microscope and is very time-consuming and prone to human error. This paper presents an automatic trap survey. For this purpose, a device has been designed that automatically scans the slat obtaining different overlapping photos. Subsequently, the images are analyzed by a Mask-RCNN neural network that segments the eggs and classifies them into 2 classes: full or hatch

2405.20700

Gecheng Chen

Gecheng Chen, Zeyu Yang, Chengwen Luo, Jianqiang Li

Self-degraded contrastive domain adaptation for industrial fault diagnosis with bi-imbalanced data

cs.AI

http://creativecommons.org/licenses/by/4.0/

Modern industrial fault diagnosis tasks often face the combined challenge of distribution discrepancy and bi-imbalance. Existing domain adaptation approaches pay little attention to the prevailing bi-imbalance, leading to poor domain adaptation performance or even negative transfer. In this work, we propose a self-degraded contrastive domain adaptation (Sd-CDA) diagnosis framework to handle the domain discrepancy under the bi-imbalanced data. It first pre-trains the feature extractor via imbalance-aware contrastive learning based on model pruning to learn the feature representation efficiently in a self-supervised manner. Then it forces the samples away from the domain boundary based on supervised contrastive domain adversarial learning (SupCon-DA) and ensures the features generated by the feature extractor are discriminative enough. Furthermore, we propose the pruned contrastive domain adversarial learning (PSupCon-DA) to pay automatically re-weighted attention to the minorities to enhance the performance towards bi-imbalanced data. We show the superiority of the proposed method via two experiments.

2405.20705

S\"oren Schleibaum

S\"oren Schleibaum, Lu Feng, Sarit Kraus, J\"org P. M\"uller

ADESSE: Advice Explanations in Complex Repeated Decision-Making Environments

cs.AI

http://creativecommons.org/licenses/by/4.0/

In the evolving landscape of human-centered AI, fostering a synergistic relationship between humans and AI agents in decision-making processes stands as a paramount challenge. This work considers a problem setup where an intelligent agent comprising a neural network-based prediction component and a deep reinforcement learning component provides advice to a human decision-maker in complex repeated decision-making environments. Whether the human decision-maker would follow the agent's advice depends on their beliefs and trust in the agent and on their understanding of the advice itself. To this end, we developed an approach named ADESSE to generate explanations about the adviser agent to improve human trust and decision-making. Computational experiments on a range of environments with varying model sizes demonstrate the applicability and scalability of ADESSE. Furthermore, an interactive game-based user study shows that participants were significantly more satisfied, achieved a higher reward in the game, and took less time to select an action when presented with explanations generated by ADESSE. These findings illuminate the critical role of tailored, human-centered explanations in AI-assisted decision-making.

2405.20978

Felton Fang

Feiteng Fang, Yuelin Bai, Shiwen Ni, Min Yang, Xiaojun Chen and Ruifeng Xu

Enhancing Noise Robustness of Retrieval-Augmented Language Models with Adaptive Adversarial Training

ACL 2024, Main Conference

cs.AI

http://creativecommons.org/licenses/by/4.0/

Large Language Models (LLMs) exhibit substantial capabilities yet encounter challenges, including hallucination, outdated knowledge, and untraceable reasoning processes. Retrieval-augmented generation (RAG) has emerged as a promising solution, integrating knowledge from external databases to mitigate these challenges. However, inappropriate retrieved passages can potentially hinder the LLMs' capacity to generate comprehensive and high-quality responses. Prior RAG studies on the robustness of retrieval noises often confine themselves to a limited set of noise types, deviating from real-world retrieval environments and limiting practical applicability. In this study, we initially investigate retrieval noises and categorize them into three distinct types, reflecting real-world environments. We analyze the impact of these various retrieval noises on the robustness of LLMs. Subsequently, we propose a novel RAG approach known as Retrieval-augmented Adaptive Adversarial Training (RAAT). RAAT leverages adaptive adversarial training to dynamically adjust the model's training process in response to retrieval noises. Concurrently, it employs multi-task learning to ensure the model's capacity to internally recognize noisy contexts. Extensive experiments demonstrate that the LLaMA-2 7B model trained using RAAT exhibits significant improvements in F1 and EM scores under diverse noise conditions. For reproducibility, we release our code and data at: https://github.com/calubkk/RAAT.

2405.21030

Benjamin Levinstein

Daniel A. Herrmann and Benjamin A. Levinstein

Standards for Belief Representations in LLMs

cs.AI

http://creativecommons.org/licenses/by/4.0/

As large language models (LLMs) continue to demonstrate remarkable abilities across various domains, computer scientists are developing methods to understand their cognitive processes, particularly concerning how (and if) LLMs internally represent their beliefs about the world. However, this field currently lacks a unified theoretical foundation to underpin the study of belief in LLMs. This article begins filling this gap by proposing adequacy conditions for a representation in an LLM to count as belief-like. We argue that, while the project of belief measurement in LLMs shares striking features with belief measurement as carried out in decision theory and formal epistemology, it also differs in ways that should change how we measure belief. Thus, drawing from insights in philosophy and contemporary practices of machine learning, we establish four criteria that balance theoretical considerations with practical constraints. Our proposed criteria include accuracy, coherence, uniformity, and use, which together help lay the groundwork for a comprehensive understanding of belief representation in LLMs. We draw on empirical work showing the limitations of using various criteria in isolation to identify belief representations.

2406.00216

Michail Mamalakis Dr

Michail Mamalakis, H\'elo\"ise de Vareilles, Graham Murray, Pietro Lio, John Suckling

The Explanation Necessity for Healthcare AI

cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

Explainability is often critical to the acceptable implementation of artificial intelligence (AI). Nowhere is this more important than healthcare where decision-making directly impacts patients and trust in AI systems is essential. This trust is often built on the explanations and interpretations the AI provides. Despite significant advancements in AI interpretability, there remains the need for clear guidelines on when and to what extent explanations are necessary in the medical context. We propose a novel categorization system with four distinct classes of explanation necessity, guiding the level of explanation required: patient or sample (local) level, cohort or dataset (global) level, or both levels. We introduce a mathematical formulation that distinguishes these categories and offers a practical framework for researchers to determine the necessity and depth of explanations required in medical AI applications. Three key factors are considered: the robustness of the evaluation protocol, the variability of expert observations, and the representation dimensionality of the application. In this perspective, we address the question: When does an AI medical application need to be explained, and at what level of detail?

2406.00392

Jonathan Cook

Jonathan Cook, Chris Lu, Edward Hughes, Joel Z. Leibo, Jakob Foerster

Artificial Generational Intelligence: Cultural Accumulation in Reinforcement Learning

cs.AI

http://creativecommons.org/licenses/by/4.0/

Cultural accumulation drives the open-ended and diverse progress in capabilities spanning human history. It builds an expanding body of knowledge and skills by combining individual exploration with inter-generational information transmission. Despite its widespread success among humans, the capacity for artificial learning agents to accumulate culture remains under-explored. In particular, approaches to reinforcement learning typically strive for improvements over only a single lifetime. Generational algorithms that do exist fail to capture the open-ended, emergent nature of cultural accumulation, which allows individuals to trade-off innovation and imitation. Building on the previously demonstrated ability for reinforcement learning agents to perform social learning, we find that training setups which balance this with independent learning give rise to cultural accumulation. These accumulating agents outperform those trained for a single lifetime with the same cumulative experience. We explore this accumulation by constructing two models under two distinct notions of a generation: episodic generations, in which accumulation occurs via in-context learning and train-time generations, in which accumulation occurs via in-weights learning. In-context and in-weights cultural accumulation can be interpreted as analogous to knowledge and skill accumulation, respectively. To the best of our knowledge, this work is the first to present general models that achieve emergent cultural accumulation in reinforcement learning, opening up new avenues towards more open-ended learning systems, as well as presenting new opportunities for modelling human culture.

2406.00415

Xuan Wu

Xuan Wu, Di Wang, Lijie Wen, Yubin Xiao, Chunguo Wu, Yuesong Wu, Chaoyu Yu, Douglas L. Maskell, and You Zhou

Neural Combinatorial Optimization Algorithms for Solving Vehicle Routing Problems: A Comprehensive Survey with Perspectives

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Although several surveys on Neural Combinatorial Optimization (NCO) solvers specifically designed to solve Vehicle Routing Problems (VRPs) have been conducted. These existing surveys did not cover the state-of-the-art (SOTA) NCO solvers emerged recently. More importantly, to provide a comprehensive taxonomy of NCO solvers with up-to-date coverage, based on our thorough review of relevant publications and preprints, we divide all NCO solvers into four distinct categories, namely Learning to Construct, Learning to Improve, Learning to Predict-Once, and Learning to Predict-Multiplicity solvers. Subsequently, we present the inadequacies of the SOTA solvers, including poor generalization, incapability to solve large-scale VRPs, inability to address most types of VRP variants simultaneously, and difficulty in comparing these NCO solvers with the conventional Operations Research algorithms. Simultaneously, we propose promising and viable directions to overcome these inadequacies. In addition, we compare the performance of representative NCO solvers from the Reinforcement, Supervised, and Unsupervised Learning paradigms across both small- and large-scale VRPs. Finally, following the proposed taxonomy, we provide an accompanying web page as a live repository for NCO solvers. Through this survey and the live repository, we hope to make the research community of NCO solvers for VRPs more thriving.

2406.00537

Lucas Vieira

Lucas Valadares Vieira, Mara Abel, Fabricio Henrique Rodrigues, Tiago Prince Sales, Claudenir M. Fonseca

Towards an ontology of portions of matter to support multi-scale analysis and provenance tracking

cs.AI

http://creativecommons.org/licenses/by/4.0/

This paper presents an ontology of portions of matter with practical implications across scientific and industrial domains. The ontology is developed under the Unified Foundational Ontology (UFO), which uses the concept of quantity to represent topologically maximally self-connected portions of matter. The proposed ontology introduces the granuleOf parthood relation, holding between objects and portions of matter. It also discusses the constitution of quantities by collections of granules, the representation of sub-portions of matter, and the tracking of matter provenance between quantities using historical relations. Lastly, a case study is presented to demonstrate the use of the portion of matter ontology in the geology domain for an Oil & Gas industry application. In the case study, we model how to represent the historical relation between an original portion of rock and the sub-portions created during the industrial process. Lastly, future research directions are outlined, including investigating granularity levels and defining a taxonomy of events.

2406.01131

Jan Deriu

Pius von D\"aniken, Jan Deriu, Don Tuggener, Mark Cieliebak

Favi-Score: A Measure for Favoritism in Automated Preference Ratings for Generative AI Evaluation

Accepted at ACL Main Conference

cs.AI

http://creativecommons.org/licenses/by/4.0/

Generative AI systems have become ubiquitous for all kinds of modalities, which makes the issue of the evaluation of such models more pressing. One popular approach is preference ratings, where the generated outputs of different systems are shown to evaluators who choose their preferences. In recent years the field shifted towards the development of automated (trained) metrics to assess generated outputs, which can be used to create preference ratings automatically. In this work, we investigate the evaluation of the metrics themselves, which currently rely on measuring the correlation to human judgments or computing sign accuracy scores. These measures only assess how well the metric agrees with the human ratings. However, our research shows that this does not tell the whole story. Most metrics exhibit a disagreement with human system assessments which is often skewed in favor of particular text generation systems, exposing a degree of favoritism in automated metrics. This paper introduces a formal definition of favoritism in preference metrics, and derives the Favi-Score, which measures this phenomenon. In particular we show that favoritism is strongly related to errors in final system rankings. Thus, we propose that preference-based metrics ought to be evaluated on both sign accuracy scores and favoritism.

2406.01139

Thomas Bolander

Thomas Bolander, Alessandro Burigana, Marco Montali

Depth-Bounded Epistemic Planning

cs.AI

http://creativecommons.org/licenses/by/4.0/

In this paper, we propose a novel algorithm for epistemic planning based on dynamic epistemic logic (DEL). The novelty is that we limit the depth of reasoning of the planning agent to an upper bound b, meaning that the planning agent can only reason about higher-order knowledge to at most (modal) depth b. The algorithm makes use of a novel type of canonical b-bisimulation contraction guaranteeing unique minimal models with respect to b-bisimulation. We show our depth-bounded planning algorithm to be sound. Additionally, we show it to be complete with respect to planning tasks having a solution within bound b of reasoning depth (and hence the iterative bound-deepening variant is complete in the standard sense). For bound b of reasoning depth, the algorithm is shown to be (b + 1)-EXPTIME complete, and furthermore fixed-parameter tractable in the number of agents and atoms. We present both a tree search and a graph search variant of the algorithm, and we benchmark an implementation of the tree search version against a baseline epistemic planner.

2406.01140

Qinggang Zhang

Qinggang Zhang, Keyu Duan, Junnan Dong, Pai Zheng, Xiao Huang

Logical Reasoning with Relation Network for Inductive Knowledge Graph Completion

cs.AI

http://creativecommons.org/publicdomain/zero/1.0/

Inductive knowledge graph completion (KGC) aims to infer the missing relation for a set of newly-coming entities that never appeared in the training set. Such a setting is more in line with reality, as real-world KGs are constantly evolving and introducing new knowledge. Recent studies have shown promising results using message passing over subgraphs to embed newly-coming entities for inductive KGC. However, the inductive capability of these methods is usually limited by two key issues. (i) KGC always suffers from data sparsity, and the situation is even exacerbated in inductive KGC where new entities often have few or no connections to the original KG. (ii) Cold-start problem. It is over coarse-grained for accurate KG reasoning to generate representations for new entities by gathering the local information from few neighbors. To this end, we propose a novel iNfOmax RelAtion Network, namely NORAN, for inductive KG completion. It aims to mine latent relation patterns for inductive KG completion. Specifically, by centering on relations, NORAN provides a hyper view towards KG modeling, where the correlations between relations can be naturally captured as entity-independent logical evidence to conduct inductive KGC. Extensive experiment results on five benchmarks show that our framework substantially outperforms the state-of-the-art KGC methods.

2406.01377

Weihao Zeng

Weihao Zeng, Joseph Campbell, Simon Stepputtis, Katia Sycara

Multi-Agent Transfer Learning via Temporal Contrastive Learning

6 pages, 6 figures

2024 IEEE International Conference on Robotics and Automation (ICRA) 2024

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper introduces a novel transfer learning framework for deep multi-agent reinforcement learning. The approach automatically combines goal-conditioned policies with temporal contrastive learning to discover meaningful sub-goals. The approach involves pre-training a goal-conditioned agent, finetuning it on the target domain, and using contrastive learning to construct a planning graph that guides the agent via sub-goals. Experiments on multi-agent coordination Overcooked tasks demonstrate improved sample efficiency, the ability to solve sparse-reward and long-horizon problems, and enhanced interpretability compared to baselines. The results highlight the effectiveness of integrating goal-conditioned policies with unsupervised temporal abstraction learning for complex multi-agent transfer learning. Compared to state-of-the-art baselines, our method achieves the same or better performances while requiring only 21.7% of the training samples.

2406.01759

Christoph Wehner

Christoph Wehner and Chrysa Iliopoulou and Tarek R. Besold

From Latent to Lucid: Transforming Knowledge Graph Embeddings into Interpretable Structures

cs.AI

http://creativecommons.org/licenses/by/4.0/

This paper introduces a post-hoc explainable AI method tailored for Knowledge Graph Embedding models. These models are essential to Knowledge Graph Completion yet criticized for their opaque, black-box nature. Despite their significant success in capturing the semantics of knowledge graphs through high-dimensional latent representations, their inherent complexity poses substantial challenges to explainability. Unlike existing methods, our approach directly decodes the latent representations encoded by Knowledge Graph Embedding models, leveraging the principle that similar embeddings reflect similar behaviors within the Knowledge Graph. By identifying distinct structures within the subgraph neighborhoods of similarly embedded entities, our method identifies the statistical regularities on which the models rely and translates these insights into human-understandable symbolic rules and facts. This bridges the gap between the abstract representations of Knowledge Graph Embedding models and their predictive outputs, offering clear, interpretable insights. Key contributions include a novel post-hoc explainable AI method for Knowledge Graph Embedding models that provides immediate, faithful explanations without retraining, facilitating real-time application even on large-scale knowledge graphs. The method's flexibility enables the generation of rule-based, instance-based, and analogy-based explanations, meeting diverse user needs. Extensive evaluations show our approach's effectiveness in delivering faithful and well-localized explanations, enhancing the transparency and trustworthiness of Knowledge Graph Embedding models.

2406.02103

Nir Greshler

Nir Greshler, David Ben Eli, Carmel Rabinovitz, Gabi Guetta, Liran Gispan, Guy Zohar, Aviv Tamar

A Bayesian Approach to Online Planning

cs.AI

http://creativecommons.org/licenses/by/4.0/

The combination of Monte Carlo tree search and neural networks has revolutionized online planning. As neural network approximations are often imperfect, we ask whether uncertainty estimates about the network outputs could be used to improve planning. We develop a Bayesian planning approach that facilitates such uncertainty quantification, inspired by classical ideas from the meta-reasoning literature. We propose a Thompson sampling based algorithm for searching the tree of possible actions, for which we prove the first (to our knowledge) finite time Bayesian regret bound, and propose an efficient implementation for a restricted family of posterior distributions. In addition we propose a variant of the Bayes-UCB method applied to trees. Empirically, we demonstrate that on the ProcGen Maze and Leaper environments, when the uncertainty estimates are accurate but the neural network output is inaccurate, our Bayesian approach searches the tree much more effectively. In addition, we investigate whether popular uncertainty estimation methods are accurate enough to yield significant gains in planning. Our code is available at: https://github.com/nirgreshler/bayesian-online-planning.

2406.02205

Jiapu Wang

Kai Sun, Jiapu Wang, Huajie Jiang, Yongli Hu, Baocai Yin

Query-Enhanced Adaptive Semantic Path Reasoning for Inductive Knowledge Graph Completion

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Conventional Knowledge graph completion (KGC) methods aim to infer missing information in incomplete Knowledge Graphs (KGs) by leveraging existing information, which struggle to perform effectively in scenarios involving emerging entities. Inductive KGC methods can handle the emerging entities and relations in KGs, offering greater dynamic adaptability. While existing inductive KGC methods have achieved some success, they also face challenges, such as susceptibility to noisy structural information during reasoning and difficulty in capturing long-range dependencies in reasoning paths. To address these challenges, this paper proposes the Query-Enhanced Adaptive Semantic Path Reasoning (QASPR) framework, which simultaneously captures both the structural and semantic information of KGs to enhance the inductive KGC task. Specifically, the proposed QASPR employs a query-dependent masking module to adaptively mask noisy structural information while retaining important information closely related to the targets. Additionally, QASPR introduces a global semantic scoring module that evaluates both the individual contributions and the collective impact of nodes along the reasoning path within KGs. The experimental results demonstrate that QASPR achieves state-of-the-art performance.

2406.02235

Tuan Dam

Tuan Dam and Odalric-Ambrym Maillard and Emilie Kaufmann

Power Mean Estimation in Stochastic Monte-Carlo Tree_Search

UAI 2024 conference

cs.AI

http://creativecommons.org/licenses/by/4.0/

Monte-Carlo Tree Search (MCTS) is a widely-used strategy for online planning that combines Monte-Carlo sampling with forward tree search. Its success relies on the Upper Confidence bound for Trees (UCT) algorithm, an extension of the UCB method for multi-arm bandits. However, the theoretical foundation of UCT is incomplete due to an error in the logarithmic bonus term for action selection, leading to the development of Fixed-Depth-MCTS with a polynomial exploration bonus to balance exploration and exploitation~\citep{shah2022journal}. Both UCT and Fixed-Depth-MCTS suffer from biased value estimation: the weighted sum underestimates the optimal value, while the maximum valuation overestimates it~\citep{coulom2006efficient}. The power mean estimator offers a balanced solution, lying between the average and maximum values. Power-UCT~\citep{dam2019generalized} incorporates this estimator for more accurate value estimates but its theoretical analysis remains incomplete. This paper introduces Stochastic-Power-UCT, an MCTS algorithm using the power mean estimator and tailored for stochastic MDPs. We analyze its polynomial convergence in estimating root node values and show that it shares the same convergence rate of $\mathcal{O}(n^{-1/2})$, with $n$ is the number of visited trajectories, as Fixed-Depth-MCTS, with the latter being a special case of the former. Our theoretical results are validated with empirical tests across various stochastic MDP environments.

2406.02723

Shiqi Zhang

Shiqi Zhang, Darshan Gadginmath, Fabio Pasqualetti

Predicting AI Agent Behavior through Approximation of the Perron-Frobenius Operator

12 pages, 4 figures, conference

cs.AI

http://creativecommons.org/licenses/by/4.0/

Predicting the behavior of AI-driven agents is particularly challenging without a preexisting model. In our paper, we address this by treating AI agents as nonlinear dynamical systems and adopting a probabilistic perspective to predict their statistical behavior using the Perron-Frobenius (PF) operator. We formulate the approximation of the PF operator as an entropy minimization problem, which can be solved by leveraging the Markovian property of the operator and decomposing its spectrum. Our data-driven methodology simultaneously approximates the PF operator to perform prediction of the evolution of the agents and also predicts the terminal probability density of AI agents, such as robotic systems and generative models. We demonstrate the effectiveness of our prediction model through extensive experiments on practical systems driven by AI algorithms.

2406.03000

Yaacov Pariente

Yaacov Pariente, Vadim Indelman

Simplification of Risk Averse POMDPs with Performance Guarantees

cs.AI

http://creativecommons.org/licenses/by/4.0/

Risk averse decision making under uncertainty in partially observable domains is a fundamental problem in AI and essential for reliable autonomous agents. In our case, the problem is modeled using partially observable Markov decision processes (POMDPs), when the value function is the conditional value at risk (CVaR) of the return. Calculating an optimal solution for POMDPs is computationally intractable in general. In this work we develop a simplification framework to speedup the evaluation of the value function, while providing performance guarantees. We consider as simplification a computationally cheaper belief-MDP transition model, that can correspond, e.g., to cheaper observation or transition models. Our contributions include general bounds for CVaR that allow bounding the CVaR of a random variable X, using a random variable Y, by assuming bounds between their cumulative distributions. We then derive bounds for the CVaR value function in a POMDP setting, and show how to bound the value function using the computationally cheaper belief-MDP transition model and without accessing the computationally expensive model in real-time. Then, we provide theoretical performance guarantees for the estimated bounds. Our results apply for a general simplification of a belief-MDP transition model and support simplification of both the observation and state transition models simultaneously.

2406.03069

Muhan Hou

Muhan Hou, Koen Hindriks, A.E. Eiben, Kim Baraka

"Give Me an Example Like This": Episodic Active Reinforcement Learning from Demonstrations

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Reinforcement Learning (RL) has achieved great success in sequential decision-making problems, but often at the cost of a large number of agent-environment interactions. To improve sample efficiency, methods like Reinforcement Learning from Expert Demonstrations (RLED) introduce external expert demonstrations to facilitate agent exploration during the learning process. In practice, these demonstrations, which are often collected from human users, are costly and hence often constrained to a limited amount. How to select the best set of human demonstrations that is most beneficial for learning therefore becomes a major concern. This paper presents EARLY (Episodic Active Learning from demonstration querY), an algorithm that enables a learning agent to generate optimized queries of expert demonstrations in a trajectory-based feature space. Based on a trajectory-level estimate of uncertainty in the agent's current policy, EARLY determines the optimized timing and content for feature-based queries. By querying episodic demonstrations as opposed to isolated state-action pairs, EARLY improves the human teaching experience and achieves better learning performance. We validate the effectiveness of our method in three simulated navigation tasks of increasing difficulty. The results show that our method is able to achieve expert-level performance for all three tasks with convergence over 30\% faster than other baseline methods when demonstrations are generated by simulated oracle policies. The results of a follow-up pilot user study (N=18) further validate that our method can still maintain a significantly better convergence in the case of human expert demonstrators while achieving a better user experience in perceived task load and consuming significantly less human time.

2406.03091

Sabah Binte Noor

Sabah Binte Noor and Fazlul Hasan Siddiqui

Improving Plan Execution Flexibility using Block-Substitution

cs.AI

http://creativecommons.org/licenses/by/4.0/

Partial-order plans in AI planning facilitate execution flexibility due to their less-constrained nature. Maximizing plan flexibility has been studied through the notions of plan deordering, and plan reordering. Plan deordering removes unnecessary action orderings within a plan, while plan reordering modifies them arbitrarily to minimize action orderings. This study, in contrast with traditional plan deordering and reordering strategies, improves a plan's flexibility by substituting its subplans with actions outside the plan for a planning problem. We exploit block deordering, which eliminates orderings in a POP by encapsulating coherent actions in blocks, to construct action blocks as candidate subplans for substitutions. In addition, this paper introduces a pruning technique for eliminating redundant actions within a BDPO plan. We also evaluate our approach when combined with MaxSAT-based reorderings. Our experimental result demonstrates a significant improvement in plan execution flexibility on the benchmark problems from International Planning Competitions (IPC), maintaining good coverage and execution time.

2406.03292

Giuseppe Primiero

Greta Coraglia and Francesco A. Genco and Pellegrino Piantadosi and Enrico Bagli and Pietro Giuffrida and Davide Posillipo and Giuseppe Primiero

Evaluating AI fairness in credit scoring with the BRIO tool

cs.AI

http://creativecommons.org/licenses/by/4.0/

We present a method for quantitative, in-depth analyses of fairness issues in AI systems with an application to credit scoring. To this aim we use BRIO, a tool for the evaluation of AI systems with respect to social unfairness and, more in general, ethically undesirable behaviours. It features a model-agnostic bias detection module, presented in \cite{DBLP:conf/beware/CoragliaDGGPPQ23}, to which a full-fledged unfairness risk evaluation module is added. As a case study, we focus on the context of credit scoring, analysing the UCI German Credit Dataset \cite{misc_statlog_(german_credit_data)_144}. We apply the BRIO fairness metrics to several, socially sensitive attributes featured in the German Credit Dataset, quantifying fairness across various demographic segments, with the aim of identifying potential sources of bias and discrimination in a credit scoring model. We conclude by combining our results with a revenue analysis.

2406.03367

Yangfan Wu

Xinrui Lin, Yangfan Wu, Huanyu Yang, Yu Zhang, Yanyong Zhang, Jianmin Ji

CLMASP: Coupling Large Language Models with Answer Set Programming for Robotic Task Planning

cs.AI

http://creativecommons.org/licenses/by-nc-sa/4.0/

Large Language Models (LLMs) possess extensive foundational knowledge and moderate reasoning abilities, making them suitable for general task planning in open-world scenarios. However, it is challenging to ground a LLM-generated plan to be executable for the specified robot with certain restrictions. This paper introduces CLMASP, an approach that couples LLMs with Answer Set Programming (ASP) to overcome the limitations, where ASP is a non-monotonic logic programming formalism renowned for its capacity to represent and reason about a robot's action knowledge. CLMASP initiates with a LLM generating a basic skeleton plan, which is subsequently tailored to the specific scenario using a vector database. This plan is then refined by an ASP program with a robot's action knowledge, which integrates implementation details into the skeleton, grounding the LLM's abstract outputs in practical robot contexts. Our experiments conducted on the VirtualHome platform demonstrate CLMASP's efficacy. Compared to the baseline executable rate of under 2% with LLM approaches, CLMASP significantly improves this to over 90%.

2406.03501

Roman Slowinski Prof.

Salvatore Greco and Roman S{\l}owi\'nski

Representation of preferences for multiple criteria decision aiding in a new seven-valued logic

cs.AI

http://creativecommons.org/licenses/by/4.0/

The seven-valued logic considered in this paper naturally arises within the rough set framework, allowing to distinguish vagueness due to imprecision from ambiguity due to coarseness. Recently, we discussed its utility for reasoning about data describing multi-attribute classification of objects. We also showed that this logic contains, as a particular case, the celebrated Belnap four-valued logic. Here, we present how the seven-valued logic, as well as the other logics that derive from it, can be used to represent preferences in the domain of Multiple Criteria Decision Aiding (MCDA). In particular, we propose new forms of outranking and value function preference models that aggregate multiple criteria taking into account imperfect preference information. We demonstrate that our approach effectively addresses common challenges in preference modeling for MCDA, such as uncertainty, imprecision, and ill-determination of performances and preferences. To this end, we present a specific procedure to construct a seven-valued preference relation and use it to define recommendations that consider robustness concerns by utilizing multiple outranking or value functions representing the decision maker s preferences. Moreover, we discuss the main properties of the proposed seven-valued preference structure and compare it with current approaches in MCDA, such as ordinal regression, robust ordinal regression, stochastic multiattribute acceptability analysis, stochastic ordinal regression, and so on. We illustrate and discuss the application of our approach using a didactic example. Finally, we propose directions for future research and potential applications of the proposed methodology.

2406.04028

Yoann Poupart

Yoann Poupart

Contrastive Sparse Autoencoders for Interpreting Planning of Chess-Playing Agents

Worskhop on Interpretable Policies in Reinforcement Learning @ RLC-2024, 18 pages and 15 figures

cs.AI

http://creativecommons.org/licenses/by/4.0/

AI led chess systems to a superhuman level, yet these systems heavily rely on black-box algorithms. This is unsustainable in ensuring transparency to the end-user, particularly when these systems are responsible for sensitive decision-making. Recent interpretability work has shown that the inner representations of Deep Neural Networks (DNNs) were fathomable and contained human-understandable concepts. Yet, these methods are seldom contextualised and are often based on a single hidden state, which makes them unable to interpret multi-step reasoning, e.g. planning. In this respect, we propose contrastive sparse autoencoders (CSAE), a novel framework for studying pairs of game trajectories. Using CSAE, we are able to extract and interpret concepts that are meaningful to the chess-agent plans. We primarily focused on a qualitative analysis of the CSAE features before proposing an automated feature taxonomy. Furthermore, to evaluate the quality of our trained CSAE, we devise sanity checks to wave spurious correlations in our results.

2406.04082

Lovis Heindrich

Lovis Heindrich, Falk Lieder

Leveraging automatic strategy discovery to teach people how to select better projects

cs.AI

http://creativecommons.org/licenses/by/4.0/

The decisions of individuals and organizations are often suboptimal because normative decision strategies are too demanding in the real world. Recent work suggests that some errors can be prevented by leveraging artificial intelligence to discover and teach prescriptive decision strategies that take people's constraints into account. So far, this line of research has been limited to simplified decision problems. This article is the first to extend this approach to a real-world decision problem, namely project selection. We develop a computational method (MGPS) that automatically discovers project selection strategies that are optimized for real people and develop an intelligent tutor that teaches the discovered strategies. We evaluated MGPS on a computational benchmark and tested the intelligent tutor in a training experiment with two control conditions. MGPS outperformed a state-of-the-art method and was more computationally efficient. Moreover, the intelligent tutor significantly improved people's decision strategies. Our results indicate that our method can improve human decision-making in naturalistic settings similar to real-world project selection, a first step towards applying strategy discovery to the real world.

cs/0002002

Miroslaw Truszczynski

Marc Denecker, Victor W. Marek, Miroslaw Truszczynski

Uniform semantic treatment of default and autoepistemic logics

Proceedings of the Seventh International Conference on Principles of Knowledge Representation and Reasoning (KR2000); 11 pages

Artificial Intelligence Journal, 143 (2003), pp. 79--122

cs.AI

We revisit the issue of connections between two leading formalisms in nonmonotonic reasoning: autoepistemic logic and default logic. For each logic we develop a comprehensive semantic framework based on the notion of a belief pair. The set of all belief pairs together with the so called knowledge ordering forms a complete lattice. For each logic, we introduce several semantics by means of fixpoints of operators on the lattice of belief pairs. Our results elucidate an underlying isomorphism of the respective semantic constructions. In particular, we show that the interpretation of defaults as modal formulas proposed by Konolige allows us to represent all semantics for default logic in terms of the corresponding semantics for autoepistemic logic. Thus, our results conclusively establish that default logic can indeed be viewed as a fragment of autoepistemic logic. However, as we also demonstrate, the semantics of Moore and Reiter are given by different operators and occupy different locations in their corresponding families of semantics. This result explains the source of the longstanding difficulty to formally relate these two semantics. In the paper, we also discuss approximating skeptical reasoning with autoepistemic and default logics and establish constructive principles behind such approximations.

cs/0002003

Miroslaw Truszczynski

Deborah East, Miroslaw Truszczynski

On the accuracy and running time of GSAT

Proceedings of the 9th Portuguese Conference on Artificial Intelligence (EPIA'99), Lecture Notes in Artificial Intelligence, vol. 1695, Springer-Verlag, 1999

cs.AI

Randomized algorithms for deciding satisfiability were shown to be effective in solving problems with thousands of variables. However, these algorithms are not complete. That is, they provide no guarantee that a satisfying assignment, if one exists, will be found. Thus, when studying randomized algorithms, there are two important characteristics that need to be considered: the running time and, even more importantly, the accuracy --- a measure of likelihood that a satisfying assignment will be found, provided one exists. In fact, we argue that without a reference to the accuracy, the notion of the running time for randomized algorithms is not well-defined. In this paper, we introduce a formal notion of accuracy. We use it to define a concept of the running time. We use both notions to study the random walk strategy GSAT algorithm. We investigate the dependence of accuracy on properties of input formulas such as clause-to-variable ratio and the number of satisfying assignments. We demonstrate that the running time of GSAT grows exponentially in the number of variables of the input formula for randomly generated 3-CNF formulas and for the formulas encoding 3- and 4-colorability of graphs.

cs/0002009

Luis Rocha

Luis M. Rocha

Syntactic Autonomy: Why There is no Autonomy without Symbols and How Self-Organization Might Evolve Them

cs.AI

Two different types of agency are discussed based on dynamically coherent and incoherent couplings with an environment respectively. I propose that until a private syntax (syntactic autonomy) is discovered by dynamically coherent agents, there are no significant or interesting types of closure or autonomy. When syntactic autonomy is established, then, because of a process of description-based selected self-organization, open-ended evolution is enabled. At this stage, agents depend, in addition to dynamics, on localized, symbolic memory, thus adding a level of dynamical incoherence to their interaction with the environment. Furthermore, it is the appearance of syntactic autonomy which enables much more interesting types of closures amongst agents which share the same syntax. To investigate how we can study the emergence of syntax from dynamical systems, experiments with cellular automata leading to emergent computation to solve non-trivial tasks are discussed. RNA editing is also mentioned as a process that may have been used to obtain a primordial biological code necessary open-ended evolution.

cs/0003008

Ken Satoh

Ken Satoh

Consistency Management of Normal Logic Program by Top-down Abductive Proof Procedure

cs.AI

This paper presents a method of computing a revision of a function-free normal logic program. If an added rule is inconsistent with a program, that is, if it leads to a situation such that no stable model exists for a new program, then deletion and addition of rules are performed to avoid inconsistency. We specify a revision by translating a normal logic program into an abductive logic program with abducibles to represent deletion and addition of rules. To compute such deletion and addition, we propose an adaptation of our top-down abductive proof procedure to compute a relevant abducibles to an added rule. We compute a minimally revised program, by choosing a minimal set of abducibles among all the sets of abducibles computed by a top-down proof procedure.

cs/0003012

John L. Pollock

John L. Pollock

Defeasible Reasoning in OSCAR

Nonmonotonic Reasoning Workshop, 2000

cs.AI

This is a system description for the OSCAR defeasible reasoner.

cs/0003016

Daniele Theseider Dupre'

Daniele Theseider Dupre' (Dipartimento di Scienze e Tecnologie Avanzate - Universita' del Piemonte Orientale, Alessandria, Italy)

Abductive and Consistency-Based Diagnosis Revisited: a Modeling Perspective

5 pages, 8th Int. Workshop on Nonmonotonic Reasoning, 2000

cs.AI

Diagnostic reasoning has been characterized logically as consistency-based reasoning or abductive reasoning. Previous analyses in the literature have shown, on the one hand, that choosing the (in general more restrictive) abductive definition may be appropriate or not, depending on the content of the knowledge base [Console&Torasso91], and, on the other hand, that, depending on the choice of the definition the same knowledge should be expressed in different form [Poole94]. Since in Model-Based Diagnosis a major problem is finding the right way of abstracting the behavior of the system to be modeled, this paper discusses the relation between modeling, and in particular abstraction in the model, and the notion of diagnosis.

cs/0003020

Antonis Kakas

Antonis Kakas

ACLP: Integrating Abduction and Constraint Solving

6 pages

cs.AI

ACLP is a system which combines abductive reasoning and constraint solving by integrating the frameworks of Abductive Logic Programming (ALP) and Constraint Logic Programming (CLP). It forms a general high-level knowledge representation environment for abductive problems in Artificial Intelligence and other areas. In ACLP, the task of abduction is supported and enhanced by its non-trivial integration with constraint solving facilitating its application to complex problems. The ACLP system is currently implemented on top of the CLP language of ECLiPSe as a meta-interpreter exploiting its underlying constraint solver for finite domains. It has been applied to the problems of planning and scheduling in order to test its computational effectiveness compared with the direct use of the (lower level) constraint solving framework of CLP on which it is built. These experiments provide evidence that the abductive framework of ACLP does not compromise significantly the computational efficiency of the solutions. Other experiments show the natural ability of ACLP to accommodate easily and in a robust way new or changing requirements of the original problem.

cs/0003021

Samir Chopra

Samir Chopra, Konstantinos Georgatos, Rohit Parikh

Relevance Sensitive Non-Monotonic Inference on Belief Sequences

cs.AI

We present a method for relevance sensitive non-monotonic inference from belief sequences which incorporates insights pertaining to prioritized inference and relevance sensitive, inconsistency tolerant belief revision. Our model uses a finite, logically open sequence of propositional formulas as a representation for beliefs and defines a notion of inference from maxiconsistent subsets of formulas guided by two orderings: a temporal sequencing and an ordering based on relevance relations between the conclusion and formulas in the sequence. The relevance relations are ternary (using context as a parameter) as opposed to standard binary axiomatizations. The inference operation thus defined easily handles iterated revision by maintaining a revision history, blocks the derivation of inconsistent answers from a possibly inconsistent sequence and maintains the distinction between explicit and implicit beliefs. In doing so, it provides a finitely presented formalism and a plausible model of reasoning for automated agents.

cs/0003023

Thomas Lukasiewicz

Thomas Lukasiewicz

Probabilistic Default Reasoning with Conditional Constraints

8 pages; to appear in Proceedings of the Eighth International Workshop on Nonmonotonic Reasoning, Special Session on Uncertainty Frameworks in Nonmonotonic Reasoning, Breckenridge, Colorado, USA, 9-11 April 2000

cs.AI

We propose a combination of probabilistic reasoning from conditional constraints with approaches to default reasoning from conditional knowledge bases. In detail, we generalize the notions of Pearl's entailment in system Z, Lehmann's lexicographic entailment, and Geffner's conditional entailment to conditional constraints. We give some examples that show that the new notions of z-, lexicographic, and conditional entailment have similar properties like their classical counterparts. Moreover, we show that the new notions of z-, lexicographic, and conditional entailment are proper generalizations of both their classical counterparts and the classical notion of logical entailment for conditional constraints.

cs/0003024

Hans Tompits

James P. Delgrande, Torsten Schaub, Hans Tompits

A Compiler for Ordered Logic Programs

cs.AI

This paper describes a system, called PLP, for compiling ordered logic programs into standard logic programs under the answer set semantics. In an ordered logic program, rules are named by unique terms, and preferences among rules are given by a set of dedicated atoms. An ordered logic program is transformed into a second, regular, extended logic program wherein the preferences are respected, in that the answer sets obtained in the transformed theory correspond with the preferred answer sets of the original theory. Since the result of the translation is an extended logic program, existing logic programming systems can be used as underlying reasoning engine. In particular, PLP is conceived as a front-end to the logic programming systems dlv and smodels.

cs/0003027

Bert Van Nuffelen

Bert Van Nuffelen

SLDNFA-system

6 pages conference:NMR2000, special track on System descriptions and demonstration

cs.AI

The SLDNFA-system results from the LP+ project at the K.U.Leuven, which investigates logics and proof procedures for these logics for declarative knowledge representation. Within this project inductive definition logic (ID-logic) is used as representation logic. Different solvers are being developed for this logic and one of these is SLDNFA. A prototype of the system is available and used for investigating how to solve efficiently problems represented in ID-logic.

cs/0003028

Hans Tompits

James P. Delgrande, Torsten Schaub, Hans Tompits

Logic Programs with Compiled Preferences

cs.AI

We describe an approach for compiling preferences into logic programs under the answer set semantics. An ordered logic program is an extended logic program in which rules are named by unique terms, and in which preferences among rules are given by a set of dedicated atoms. An ordered logic program is transformed into a second, regular, extended logic program wherein the preferences are respected, in that the answer sets obtained in the transformed theory correspond with the preferred answer sets of the original theory. Our approach allows both the specification of static orderings (as found in most previous work), in which preferences are external to a logic program, as well as orderings on sets of rules. In large part then, we are interested in describing a general methodology for uniformly incorporating preference information in a logic program. Since the result of our translation is an extended logic program, we can make use of existing implementations, such as dlv and smodels. To this end, we have developed a compiler, available on the web, as a front-end for these programming systems.

cs/0003029

Nedra Mellouli

Nedra Mellouli, Bernadette Bouchon-Meunier

Fuzzy Approaches to Abductive Inference

7 pages and 8 files

cs.AI

This paper proposes two kinds of fuzzy abductive inference in the framework of fuzzy rule base. The abductive inference processes described here depend on the semantic of the rule. We distinguish two classes of interpretation of a fuzzy rule, certainty generation rules and possible generation rules. In this paper we present the architecture of abductive inference in the first class of interpretation. We give two kinds of problem that we can resolve by using the proposed models of inference.

cs/0003030

Bert Van Nuffelen

Bert Van Nuffelen, Marc Denecker

Problem solving in ID-logic with aggregates: some experiments

9 pages conference: NMR2000, special track on abductive reasoning

cs.AI

The goal of the LP+ project at the K.U.Leuven is to design an expressive logic, suitable for declarative knowledge representation, and to develop intelligent systems based on Logic Programming technology for solving computational problems using the declarative specifications. The ID-logic is an integration of typed classical logic and a definition logic. Different abductive solvers for this language are being developed. This paper is a report of the integration of high order aggregates into ID-logic and the consequences on the solver SLDNFA.

cs/0003031

Robert E. Mercer

Carmen Vodislav and Robert E. Mercer

Optimal Belief Revision

NMR'2000 Workshop 6 pages

cs.AI

We propose a new approach to belief revision that provides a way to change knowledge bases with a minimum of effort. We call this way of revising belief states optimal belief revision. Our revision method gives special attention to the fact that most belief revision processes are directed to a specific informational objective. This approach to belief change is founded on notions such as optimal context and accessibility. For the sentential model of belief states we provide both a formal description of contexts as sub-theories determined by three parameters and a method to construct contexts. Next, we introduce an accessibility ordering for belief sets, which we then use for selecting the best (optimal) contexts with respect to the processing effort involved in the revision. Then, for finitely axiomatizable knowledge bases, we characterize a finite accessibility ranking from which the accessibility ordering for the entire base is generated and show how to determine the ranking of an arbitrary sentence in the language. Finally, we define the adjustment of the accessibility ranking of a revised base of a belief set.

cs/0003032

Henrik Grosskreutz

Henrik Grosskreutz, Gerhard Lakemeyer

cc-Golog: Towards More Realistic Logic-Based Robot Controllers

cs.AI

High-level robot controllers in realistic domains typically deal with processes which operate concurrently, change the world continuously, and where the execution of actions is event-driven as in ``charge the batteries as soon as the voltage level is low''. While non-logic-based robot control languages are well suited to express such scenarios, they fare poorly when it comes to projecting, in a conspicuous way, how the world evolves when actions are executed. On the other hand, a logic-based control language like \congolog, based on the situation calculus, is well-suited for the latter. However, it has problems expressing event-driven behavior. In this paper, we show how these problems can be overcome by first extending the situation calculus to support continuous change and event-driven behavior and then presenting \ccgolog, a variant of \congolog which is based on the extended situation calculus. One benefit of \ccgolog is that it narrows the gap in expressiveness compared to non-logic-based control languages while preserving a semantically well-founded projection mechanism.

cs/0003033

Ilkka Niemela

Ilkka Niemela, Patrik Simons, Tommi Syrjanen

Smodels: A System for Answer Set Programming

Proceedings of the 8th International Workshop on Non-Monotonic Reasoning, April 9-11, 2000, Breckenridge, Colorado 4 pages, uses aaai.sty

cs.AI

The Smodels system implements the stable model semantics for normal logic programs. It handles a subclass of programs which contain no function symbols and are domain-restricted but supports extensions including built-in functions as well as cardinality and weight constraints. On top of this core engine more involved systems can be built. As an example, we have implemented total and partial stable model computation for disjunctive logic programs. An interesting application method is based on answer set programming, i.e., encoding an application problem as a set of rules so that its solutions are captured by the stable models of the rules. Smodels has been applied to a number of areas including planning, model checking, reachability analysis, product configuration, dynamic constraint satisfaction, and feature interaction.

cs/0003034

Francesca Toni

Antonis Kakas, Rob Miller, Francesca Toni

E-RES: A System for Reasoning about Actions, Events and Observations

Proceedings of the 8th International Workshop on Non-Monotonic Reasoning, April 9-11, 2000, Breckenridge, Colorado. 6 pages

cs.AI

E-RES is a system that implements the Language E, a logic for reasoning about narratives of action occurrences and observations. E's semantics is model-theoretic, but this implementation is based on a sound and complete reformulation of E in terms of argumentation, and uses general computational techniques of argumentation frameworks. The system derives sceptical non-monotonic consequences of a given reformulated theory which exactly correspond to consequences entailed by E's model-theory. The computation relies on a complimentary ability of the system to derive credulous non-monotonic consequences together with a set of supporting assumptions which is sufficient for the (credulous) conclusion to hold. E-RES allows theories to contain general action laws, statements about action occurrences, observations and statements of ramifications (or universal laws). It is able to derive consequences both forward and backward in time. This paper gives a short overview of the theoretical basis of E-RES and illustrates its use on a variety of examples. Currently, E-RES is being extended so that the system can be used for planning.

cs/0003037

Hans Tompits

Uwe Egly, Thomas Eiter, Hans Tompits, Stefan Woltran

QUIP - A Tool for Computing Nonmonotonic Reasoning Tasks

cs.AI

In this paper, we outline the prototype of an automated inference tool, called QUIP, which provides a uniform implementation for several nonmonotonic reasoning formalisms. The theoretical basis of QUIP is derived from well-known results about the computational complexity of nonmonotonic logics and exploits a representation of the different reasoning tasks in terms of quantified boolean formulae.

cs/0003038

Richard Watson

Richard Watson

A Splitting Set Theorem for Epistemic Specifications

To be published in Proceedings of NMR 2000 Workshop. 6 pages

cs.AI

Over the past decade a considerable amount of research has been done to expand logic programming languages to handle incomplete information. One such language is the language of epistemic specifications. As is usual with logic programming languages, the problem of answering queries is intractable in the general case. For extended disjunctive logic programs, an idea that has proven useful in simplifying the investigation of answer sets is the use of splitting sets. In this paper we will present an extended definition of splitting sets that will be applicable to epistemic specifications. Furthermore, an extension of the splitting set theorem will be presented. Also, a characterization of stratified epistemic specifications will be given in terms of splitting sets. This characterization leads us to an algorithmic method of computing world views of a subclass of epistemic logic programs.

cs/0003039

Ilkka Niemela

Maarit Hietalahti, Fabio Massacci, Ilkka Niemela

DES: a Challenge Problem for Nonmonotonic Reasoning Systems

10 pages, 1 Postscript figure, uses aaai.sty and graphicx.sty

cs.AI

The US Data Encryption Standard, DES for short, is put forward as an interesting benchmark problem for nonmonotonic reasoning systems because (i) it provides a set of test cases of industrial relevance which shares features of randomly generated problems and real-world problems, (ii) the representation of DES using normal logic programs with the stable model semantics is simple and easy to understand, and (iii) this subclass of logic programs can be seen as an interesting special case for many other formalizations of nonmonotonic reasoning. In this paper we present two encodings of DES as logic programs: a direct one out of the standard specifications and an optimized one extending the work of Massacci and Marraro. The computational properties of the encodings are studied by using them for DES key search with the Smodels system as the implementation of the stable model semantics. Results indicate that the encodings and Smodels are quite competitive: they outperform state-of-the-art SAT-checkers working with an optimized encoding of DES into SAT and are comparable with a SAT-checker that is customized and tuned for the optimized SAT encoding.

cs/0003042

Vladimir Lifschitz

Yuliya Babovich, Esra Erdem and Vladimir Lifschitz

Fages' Theorem and Answer Set Programming

cs.AI

We generalize a theorem by Francois Fages that describes the relationship between the completion semantics and the answer set semantics for logic programs with negation as failure. The study of this relationship is important in connection with the emergence of answer set programming. Whenever the two semantics are equivalent, answer sets can be computed by a satisfiability solver, and the use of answer set solvers such as smodels and dlv is unnecessary. A logic programming representation of the blocks world due to Ilkka Niemelae is discussed as an example.

cs/0003044

Adnan

Adnan Darwiche

On the tractable counting of theory models and its application to belief revision and truth maintenance

cs.AI

We introduced decomposable negation normal form (DNNF) recently as a tractable form of propositional theories, and provided a number of powerful logical operations that can be performed on it in polynomial time. We also presented an algorithm for compiling any conjunctive normal form (CNF) into DNNF and provided a structure-based guarantee on its space and time complexity. We present in this paper a linear-time algorithm for converting an ordered binary decision diagram (OBDD) representation of a propositional theory into an equivalent DNNF, showing that DNNFs scale as well as OBDDs. We also identify a subclass of DNNF which we call deterministic DNNF, d-DNNF, and show that the previous complexity guarantees on compiling DNNF continue to hold for this stricter subclass, which has stronger properties. In particular, we present a new operation on d-DNNF which allows us to count its models under the assertion, retraction and flipping of every literal by traversing the d-DNNF twice. That is, after such traversal, we can test in constant-time: the entailment of any literal by the d-DNNF, and the consistency of the d-DNNF under the retraction or flipping of any literal. We demonstrate the significance of these new operations by showing how they allow us to implement linear-time, complete truth maintenance systems and linear-time, complete belief revision systems for two important classes of propositional theories.

cs/0003047

Hans-Peter Stoerr

Steffen Hoelldobler and Hans-Peter Stoerr

BDD-based reasoning in the fluent calculus - first results

9 pages; Workshop on Nonmonotonic Reasoning 2000 (NMR 2000)

cs.AI

The paper reports on first preliminary results and insights gained in a project aiming at implementing the fluent calculus using methods and techniques based on binary decision diagrams. After reporting on an initial experiment showing promising results we discuss our findings concerning various techniques and heuristics used to speed up the reasoning process.

cs/0003049

Rob Miller

Antonis Kakas, Rob Miller and Francesca Toni

Planning with Incomplete Information

Proceedings of the 8th International Workshop on Non-Monotonic Reasoning, April 9-11, 2000, Breckenridge, Colorado

cs.AI

Planning is a natural domain of application for frameworks of reasoning about actions and change. In this paper we study how one such framework, the Language E, can form the basis for planning under (possibly) incomplete information. We define two types of plans: weak and safe plans, and propose a planner, called the E-Planner, which is often able to extend an initial weak plan into a safe plan even though the (explicit) information available is incomplete, e.g. for cases where the initial state is not completely known. The E-Planner is based upon a reformulation of the Language E in argumentation terms and a natural proof theory resulting from the reformulation. It uses an extension of this proof theory by means of abduction for the generation of plans and adopts argumentation-based techniques for extending weak plans into safe plans. We provide representative examples illustrating the behaviour of the E-Planner, in particular for cases where the status of fluents is incompletely known.

cs/0003051

Miroslaw Truszczynski

Renata Wassermann

Local Diagnosis

cs.AI

In an earlier work, we have presented operations of belief change which only affect the relevant part of a belief base. In this paper, we propose the application of the same strategy to the problem of model-based diangosis. We first isolate the subset of the system description which is relevant for a given observation and then solve the diagnosis problem for this subset.

cs/0003052

James P. Delgrande

James Delgrande and Torsten Schaub

A Consistency-Based Model for Belief Change: Preliminary Report

cs.AI

We present a general, consistency-based framework for belief change. Informally, in revising K by A, we begin with A and incorporate as much of K as consistently possible. Formally, a knowledge base K and sentence A are expressed, via renaming propositions in K, in separate languages. Using a maximization process, we assume the languages are the same insofar as consistently possible. Lastly, we express the resultant knowledge base in a single language. There may be more than one way in which A can be so extended by K: in choice revision, one such ``extension'' represents the revised state; alternately revision consists of the intersection of all such extensions. The most general formulation of our approach is flexible enough to express other approaches to revision and update, the merging of knowledge bases, and the incorporation of static and dynamic integrity constraints. Our framework differs from work based on ordinal conditional functions, notably with respect to iterated revision. We argue that the approach is well-suited for implementation: the choice revision operator gives better complexity results than general revision; the approach can be expressed in terms of a finite knowledge base; and the scope of a revision can be restricted to just those propositions mentioned in the sentence for revision A.

cs/0003059

WIlliams

Mary-Anne Williams and Aidan Sims

SATEN: An Object-Oriented Web-Based Revision and Extraction Engine

The implementation of SATEN can be found at http://cafe.newcastle.edu.au/saten

cs.AI

SATEN is an object-oriented web-based extraction and belief revision engine. It runs on any computer via a Java 1.1 enabled browser such as Netscape 4. SATEN performs belief revision based on the AGM approach. The extraction and belief revision reasoning engines operate on a user specified ranking of information. One of the features of SATEN is that it can be used to integrate mutually inconsistent commensuate rankings into a consistent ranking.

cs/0003061

Deborah East

Deborah East and Miroslaw Truszczynski

dcs: An Implementation of DATALOG with Constraints

6 pages (AAAI format), 4 ps figures; System descriptions and demonstration Session, 8th Intl. Workshop on Non-Monotonic Reasoning

cs.AI

Answer-set programming (ASP) has emerged recently as a viable programming paradigm. We describe here an ASP system, DATALOG with constraints or DC, based on non-monotonic logic. Informally, DC theories consist of propositional clauses (constraints) and of Horn rules. The semantics is a simple and natural extension of the semantics of the propositional logic. However, thanks to the presence of Horn rules in the system, modeling of transitive closure becomes straightforward. We describe the syntax, use and implementation of DC and provide experimental results.

cs/0003077

Deborah East

Deborah East and Miroslaw Truszczynski

DATALOG with constraints - an answer-set programming system

6 pages, 5 figures, will appear in Proceedings of AAAI-2000

cs.AI

Answer-set programming (ASP) has emerged recently as a viable programming paradigm well attuned to search problems in AI, constraint satisfaction and combinatorics. Propositional logic is, arguably, the simplest ASP system with an intuitive semantics supporting direct modeling of problem constraints. However, for some applications, especially those requiring that transitive closure be computed, it requires additional variables and results in large theories. Consequently, it may not be a practical computational tool for such problems. On the other hand, ASP systems based on nonmonotonic logics, such as stable logic programming, can handle transitive closure computation efficiently and, in general, yield very concise theories as problem representations. Their semantics is, however, more complex. Searching for the middle ground, in this paper we introduce a new nonmonotonic logic, DATALOG with constraints or DC. Informally, DC theories consist of propositional clauses (constraints) and of Horn rules. The semantics is a simple and natural extension of the semantics of the propositional logic. However, thanks to the presence of Horn rules in the system, modeling of transitive closure becomes straightforward. We describe the syntax and semantics of DC, and study its properties. We discuss an implementation of DC and present results of experimental study of the effectiveness of DC, comparing it with CSAT, a satisfiability checker and SMODELS implementation of stable logic programming. Our results show that DC is competitive with the other two approaches, in case of many search problems, often yielding much more efficient solutions.

cs/0003080

Krzysztof R. Apt

Krzysztof R. Apt

Some Remarks on Boolean Constraint Propagation

14 pages. To appear in: New Trends in Constraints, Papers from the Joint ERCIM/Compulog-Net Workshop Cyprus, October 25-27, 1999. Springer-Verlag Lecture Notes in Artificial Intelligence

cs.AI

We study here the well-known propagation rules for Boolean constraints. First we propose a simple notion of completeness for sets of such rules and establish a completeness result. Then we show an equivalence in an appropriate sense between Boolean constraint propagation and unit propagation, a form of resolution for propositional logic. Subsequently we characterize one set of such rules by means of the notion of hyper-arc consistency introduced in (Mohr and Masini 1988). Also, we clarify the status of a similar, though different, set of rules introduced in (Simonis 1989a) and more fully in (Codognet and Diaz 1996).

cs/0005031

Joseph Y. Halpern

Joseph Y. Halpern

Conditional Plausibility Measures and Bayesian Networks

Journal Of Artificial Intelligence Research, Volume 14, pages 359-389, 2001

10.1613/jair.817

cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

A general notion of algebraic conditional plausibility measures is defined. Probability measures, ranking functions, possibility measures, and (under the appropriate definitions) sets of probability measures can all be viewed as defining algebraic conditional plausibility measures. It is shown that algebraic conditional plausibility measures can be represented using Bayesian networks.

cs/0006043

Sylvain Piechowiak

S. Piechowiak, J. Rodriguez

Constraint compiling into rules formalism constraint compiling into rules formalism for dynamic CSPs computing

14 pages

cs.AI

In this paper we present a rule based formalism for filtering variables domains of constraints. This formalism is well adapted for solving dynamic CSP. We take diagnosis as an instance problem to illustrate the use of these rules. A diagnosis problem is seen like finding all the minimal sets of constraints to be relaxed in the constraint network that models the device to be diagnosed

cs/0007004

Alejandro Zunino

Alejandro Zunino and Analia Amandi

Brainstorm/J: a Java Framework for Intelligent Agents

15 pages. To be published in Proceedings of the Second Argentinian Symposium on Artificial Intelligence (ASAI'2000 - 29th JAIIO). September 2000. Tandil, Buenos Aires, Argentina. See http://www.exa.unicen.edu.ar/~azunino

cs.AI

Despite the effort of many researchers in the area of multi-agent systems (MAS) for designing and programming agents, a few years ago the research community began to take into account that common features among different MAS exists. Based on these common features, several tools have tackled the problem of agent development on specific application domains or specific types of agents. As a consequence, their scope is restricted to a subset of the huge application domain of MAS. In this paper we propose a generic infrastructure for programming agents whose name is Brainstorm/J. The infrastructure has been implemented as an object oriented framework. As a consequence, our approach supports a broader scope of MAS applications than previous efforts, being flexible and reusable.

cs/0010037

Umberto Straccia

Umberto Straccia

On the relationship between fuzzy logic and four-valued relevance logic

cs.AI

In fuzzy propositional logic, to a proposition a partial truth in [0,1] is assigned. It is well known that under certain circumstances, fuzzy logic collapses to classical logic. In this paper, we will show that under dual conditions, fuzzy logic collapses to four-valued (relevance) logic, where propositions have truth-value true, false, unknown, or contradiction. As a consequence, fuzzy entailment may be considered as ``in between'' four-valued (relevance) entailment and classical entailment.

cs/0011012

Joseph Y. Halpern

Joseph Y. Halpern and Judea Pearl

Causes and Explanations: A Structural-Model Approach, Part I: Causes

Part II of the paper (on Explanation) is also on the arxiv. Previously the two parts were submitted as one paper. To appear in the British Journal for the Philosophy of Science

cs.AI

We propose a new definition of actual cause, using structural equations to model counterfactuals. We show that the definition yields a plausible and elegant account of causation that handles well examples which have caused problems for other definitions and resolves major difficulties in the traditional account.

cs/0011030

Emmanuel De Mot

Nikolay Pelov, Emmanuel De Mot, Marc Denecker

Logic Programming Approaches for Representing and Solving Constraint Satisfaction Problems: A Comparison

15 pages, 3 eps-figures

LPAR 2000, Lecture Notes in Artificial Intelligence, vol. 1955, Springer, 2000, pp. 225-239

cs.AI

Many logic programming based approaches can be used to describe and solve combinatorial search problems. On the one hand there is constraint logic programming which computes a solution as an answer substitution to a query containing the variables of the constraint satisfaction problem. On the other hand there are systems based on stable model semantics, abductive systems, and first order logic model generators which compute solutions as models of some theory. This paper compares these different approaches from the point of view of knowledge representation (how declarative are the programs) and from the point of view of performance (how good are they at solving typical problems).