arXiv - CS - Symbolic Computation最新文献_第9页

Two Algorithms for Computing Rational Univariate Representations of Zero-Dimensional Ideals with Parameters 计算有参数零维理想的有理单变量表示的两种算法

arXiv - CS - Symbolic Computation Pub Date : 2024-03-25 DOI: arxiv-2403.16519

Dingkang Wang, Jingjing Wei, Fanghui Xiao, Xiaopeng Zheng

{"title":"Two Algorithms for Computing Rational Univariate Representations of Zero-Dimensional Ideals with Parameters","authors":"Dingkang Wang, Jingjing Wei, Fanghui Xiao, Xiaopeng Zheng","doi":"arxiv-2403.16519","DOIUrl":"https://doi.org/arxiv-2403.16519","url":null,"abstract":"Two algorithms for computing the rational univariate representation of\u0000zero-dimensional ideals with parameters are presented in the paper. Different\u0000from the rational univariate representation of zero-dimensional ideals without\u0000parameters, the number of zeros of zero-dimensional ideals with parameters\u0000under various specializations is different, which leads to choosing and\u0000checking the separating element, the key to computing the rational univariate\u0000representation, is difficult. In order to pick out the separating element, by\u0000partitioning the parameter space we can ensure that under each branch the ideal\u0000has the same number of zeros. Subsequently with the help of the extended\u0000subresultant theorem for parametric cases, two ideas are given to conduct the\u0000further partition of parameter space for choosing and checking the separating\u0000element. Based on these, we give two algorithms for computing rational\u0000univariate representations of zero-dimensional ideals with parameters.\u0000Furthermore, the two algorithms have been implemented on the computer algebra\u0000system Singular. Experimental data show that the second algorithm has the\u0000better performance in contrast to the first one.","PeriodicalId":501033,"journal":{"name":"arXiv - CS - Symbolic Computation","volume":"181 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140298693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

OSVAuto: semi-automatic verifier for functional specifications of operating systems OSVAuto：操作系统功能规格的半自动验证器

arXiv - CS - Symbolic Computation Pub Date : 2024-03-20 DOI: arxiv-2403.13457

Yulun Wu, Bohua Zhan, Bican Xia

引用次数: 0

Self-Attention Based Semantic Decomposition in Vector Symbolic Architectures 矢量符号架构中基于自注意力的语义分解

arXiv - CS - Symbolic Computation Pub Date : 2024-03-20 DOI: arxiv-2403.13218

Calvin Yeung, Prathyush Poduval, Mohsen Imani

{"title":"Self-Attention Based Semantic Decomposition in Vector Symbolic Architectures","authors":"Calvin Yeung, Prathyush Poduval, Mohsen Imani","doi":"arxiv-2403.13218","DOIUrl":"https://doi.org/arxiv-2403.13218","url":null,"abstract":"Vector Symbolic Architectures (VSAs) have emerged as a novel framework for\u0000enabling interpretable machine learning algorithms equipped with the ability to\u0000reason and explain their decision processes. The basic idea is to represent\u0000discrete information through high dimensional random vectors. Complex data\u0000structures can be built up with operations over vectors such as the \"binding\"\u0000operation involving element-wise vector multiplication, which associates data\u0000together. The reverse task of decomposing the associated elements is a\u0000combinatorially hard task, with an exponentially large search space. The main\u0000algorithm for performing this search is the resonator network, inspired by\u0000Hopfield network-based memory search operations. In this work, we introduce a new variant of the resonator network, based on\u0000self-attention based update rules in the iterative search problem. This update\u0000rule, based on the Hopfield network with log-sum-exp energy function and\u0000norm-bounded states, is shown to substantially improve the performance and rate\u0000of convergence. As a result, our algorithm enables a larger capacity for\u0000associative memory, enabling applications in many tasks like perception based\u0000pattern recognition, scene decomposition, and object reasoning. We substantiate\u0000our algorithm with a thorough evaluation and comparisons to baselines.","PeriodicalId":501033,"journal":{"name":"arXiv - CS - Symbolic Computation","volume":"42 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140196921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hierarchical NeuroSymbolic Approach for Action Quality Assessment 行动质量评估的层次神经符号法

arXiv - CS - Symbolic Computation Pub Date : 2024-03-20 DOI: arxiv-2403.13798

Lauren Okamoto, Paritosh Parmar

{"title":"Hierarchical NeuroSymbolic Approach for Action Quality Assessment","authors":"Lauren Okamoto, Paritosh Parmar","doi":"arxiv-2403.13798","DOIUrl":"https://doi.org/arxiv-2403.13798","url":null,"abstract":"Action quality assessment (AQA) applies computer vision to quantitatively\u0000assess the performance or execution of a human action. Current AQA approaches\u0000are end-to-end neural models, which lack transparency and tend to be biased\u0000because they are trained on subjective human judgements as ground-truth. To\u0000address these issues, we introduce a neuro-symbolic paradigm for AQA, which\u0000uses neural networks to abstract interpretable symbols from video data and\u0000makes quality assessments by applying rules to those symbols. We take diving as\u0000the case study. We found that domain experts prefer our system and find it more\u0000informative than purely neural approaches to AQA in diving. Our system also\u0000achieves state-of-the-art action recognition and temporal segmentation, and\u0000automatically generates a detailed report that breaks the dive down into its\u0000elements and provides objective scoring with visual evidence. As verified by a\u0000group of domain experts, this report may be used to assist judges in scoring,\u0000help train judges, and provide feedback to divers. We will open-source all of\u0000our annotated training data and code for ease of reproducibility.","PeriodicalId":501033,"journal":{"name":"arXiv - CS - Symbolic Computation","volume":"141 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140196983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Routing and Scheduling in Answer Set Programming applied to Multi-Agent Path Finding: Preliminary Report 应用于多代理路径查找的答案集编程中的路由和调度：初步报告

arXiv - CS - Symbolic Computation Pub Date : 2024-03-18 DOI: arxiv-2403.12153

Roland Kaminski, Torsten Schaub, Tran Cao Son, Jiří Švancara, Philipp Wanko

引用次数: 0

Reasoning Abilities of Large Language Models: In-Depth Analysis on the Abstraction and Reasoning Corpus 大型语言模型的推理能力：对抽象与推理语料库的深入分析

arXiv - CS - Symbolic Computation Pub Date : 2024-03-18 DOI: arxiv-2403.11793

Seungpil Lee, Woochang Sim, Donghyeon Shin, Sanha Hwang, Wongyu Seo, Jiwon Park, Seokki Lee, Sejin Kim, Sundong Kim

引用次数: 0

Complexity Classification of Complex-Weighted Counting Acyclic Constraint Satisfaction Problems 复杂加权计数无循环约束满足问题的复杂性分类

arXiv - CS - Symbolic Computation Pub Date : 2024-03-14 DOI: arxiv-2403.09145

Tomoyuki Yamakami

引用次数: 0

ArgMed-Agents: Explainable Clinical Decision Reasoning with Large Language Models via Argumentation Schemes ArgMed-Agents：通过论证方案利用大型语言模型进行可解释的临床决策推理

arXiv - CS - Symbolic Computation Pub Date : 2024-03-10 DOI: arxiv-2403.06294

Shengxin Hong, Liang Xiao, Xin Zhang, Jianxia Chen

{"title":"ArgMed-Agents: Explainable Clinical Decision Reasoning with Large Language Models via Argumentation Schemes","authors":"Shengxin Hong, Liang Xiao, Xin Zhang, Jianxia Chen","doi":"arxiv-2403.06294","DOIUrl":"https://doi.org/arxiv-2403.06294","url":null,"abstract":"There are two main barriers to using large language models (LLMs) in clinical\u0000reasoning. Firstly, while LLMs exhibit significant promise in Natural Language\u0000Processing (NLP) tasks, their performance in complex reasoning and planning\u0000falls short of expectations. Secondly, LLMs use uninterpretable methods to make\u0000clinical decisions that are fundamentally different from the clinician's\u0000cognitive processes. This leads to user distrust. In this paper, we present a\u0000multi-agent framework called ArgMed-Agents, which aims to enable LLM-based\u0000agents to make explainable clinical decision reasoning through interaction.\u0000ArgMed-Agents performs self-argumentation iterations via Argumentation Scheme\u0000for Clinical Decision (a reasoning mechanism for modeling cognitive processes\u0000in clinical reasoning), and then constructs the argumentation process as a\u0000directed graph representing conflicting relationships. Ultimately, Reasoner(a\u0000symbolic solver) identify a series of rational and coherent arguments to\u0000support decision. ArgMed-Agents enables LLMs to mimic the process of clinical\u0000argumentative reasoning by generating explanations of reasoning in a\u0000self-directed manner. The setup experiments show that ArgMed-Agents not only\u0000improves accuracy in complex clinical decision reasoning problems compared to\u0000other prompt methods, but more importantly, it provides users with decision\u0000explanations that increase their confidence.","PeriodicalId":501033,"journal":{"name":"arXiv - CS - Symbolic Computation","volume":"144 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140107605","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Learning Guided Automated Reasoning: A Brief Survey 学习引导式自动推理：简要调查

arXiv - CS - Symbolic Computation Pub Date : 2024-03-06 DOI: arxiv-2403.04017

Lasse Blaauwbroek, David Cerna, Thibault Gauthier, Jan Jakubův, Cezary Kaliszyk, Martin Suda, Josef Urban

{"title":"Learning Guided Automated Reasoning: A Brief Survey","authors":"Lasse Blaauwbroek, David Cerna, Thibault Gauthier, Jan Jakubův, Cezary Kaliszyk, Martin Suda, Josef Urban","doi":"arxiv-2403.04017","DOIUrl":"https://doi.org/arxiv-2403.04017","url":null,"abstract":"Automated theorem provers and formal proof assistants are general reasoning\u0000systems that are in theory capable of proving arbitrarily hard theorems, thus\u0000solving arbitrary problems reducible to mathematics and logical reasoning. In\u0000practice, such systems however face large combinatorial explosion, and\u0000therefore include many heuristics and choice points that considerably influence\u0000their performance. This is an opportunity for trained machine learning\u0000predictors, which can guide the work of such reasoning systems. Conversely,\u0000deductive search supported by the notion of logically valid proof allows one to\u0000train machine learning systems on large reasoning corpora. Such bodies of proof\u0000are usually correct by construction and when combined with more and more\u0000precise trained guidance they can be boostrapped into very large corpora, with\u0000increasingly long reasoning chains and possibly novel proof ideas. In this\u0000paper we provide an overview of several automated reasoning and theorem proving\u0000domains and the learning and AI methods that have been so far developed for\u0000them. These include premise selection, proof guidance in several settings, AI\u0000systems and feedback loops iterating between reasoning and learning, and\u0000symbolic classification problems.","PeriodicalId":501033,"journal":{"name":"arXiv - CS - Symbolic Computation","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140074007","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Saturating Sorting without Sorts 无排序饱和排序

arXiv - CS - Symbolic Computation Pub Date : 2024-03-06 DOI: arxiv-2403.03712

Pamina Georgiou, Márton Hajdu, Laura Kovács

{"title":"Saturating Sorting without Sorts","authors":"Pamina Georgiou, Márton Hajdu, Laura Kovács","doi":"arxiv-2403.03712","DOIUrl":"https://doi.org/arxiv-2403.03712","url":null,"abstract":"We present a first-order theorem proving framework for establishing the\u0000correctness of functional programs implementing sorting algorithms with\u0000recursive data structures. We formalize the semantics of recursive programs in many-sorted first-order\u0000logic and integrate sortedness/permutation properties within our first-order\u0000formalization. Rather than focusing on sorting lists of elements of specific\u0000first-order theories, such as integer arithmetic, our list formalization relies\u0000on a sort parameter abstracting (arithmetic) theories and hence concrete sorts.\u0000We formalize the permutation property of lists in first-order logic so that we\u0000automatically prove verification conditions of such algorithms purely by\u0000superpositon-based first-order reasoning. Doing so, we adjust recent efforts\u0000for automating inducion in saturation. We advocate a compositional approach for\u0000automating proofs by induction required to verify functional programs\u0000implementing and preserving sorting and permutation properties over\u0000parameterized list structures. Our work turns saturation-based first-order\u0000theorem proving into an automated verification engine by (i) guiding automated\u0000inductive reasoning with manual proof splits and (ii) fully automating\u0000inductive reasoning in saturation. We showcase the applicability of our\u0000framework over recursive sorting algorithms, including Mergesort and Quicksort.","PeriodicalId":501033,"journal":{"name":"arXiv - CS - Symbolic Computation","volume":"54 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140056986","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0