Many experts argue that the future of artificial intelligence is limited by the field's ability to integrate symbolic logical reasoning into deep learning architectures. The recently proposed differentiable MAXSAT solver, SATNet, was a breakthrough in its capacity to integrate with a traditional neural network and solve visual reasoning problems. For instance, it can learn the rules of Sudoku purely from image examples. Despite its success, SATNet was shown to succumb to a key challenge in neurosymbolic systems known as the Symbol Grounding Problem: the inability to map visual inputs to symbolic variables without explicit supervision ("label leakage"). In this work, we present a self-supervised pre-training pipeline that enables SATNet to overcome this limitation, thus broadening the class of problems that SATNet architectures can solve to include datasets where no intermediary labels are available at all. We demonstrate that our method allows SATNet to attain full accuracy even with a harder problem setup that prevents any label leakage. We additionally introduce a proofreading method that further improves the performance of SATNet architectures, beating the state-of-the-art on Visual Sudoku.
KG-TRACE: A Neuro-Symbolic Framework for Mechanistic Grounding in Antimicrobial Resistance Prediction
National Institute of Technology Kurukshetra(国立库鲁克舍特拉理工学院);Indian Institute of Information Technology Manipur(印度曼尼普尔信息技术学院);Purdue University(普渡大学);Indian Institute of Technology Indore(印度印多尔理工学院)
Symbol Grounding in Neuro-Symbolic AI: A Gentle Introduction to Reasoning Shortcuts
University of Trento(特伦托大学);Vrije Universiteit Amsterdam(阿姆斯特丹自由大学);Sapienza University of Rome(罗马大学);University of Edinburgh(爱丁堡大学);Huawei Labs(华为实验室)
BioProAgent: Neuro-Symbolic Grounding for Constrained Scientific Planning
School of AI for Science, Peking University(科学人工智能学院,北京大学);School of Electronic and Computer Engineering, Peking University(电子与计算机工程学院,北京大学);School of Computer Science, Peking University(计算机科学学院,北京大学)
PIGLeT: Language Grounding Through Neuro-Symbolic Interaction in a 3D World
Paul G. Allen School of Computer Science & Engineering, University of Washington(华盛顿大学保罗·G·艾伦计算机科学与工程学院);Allen Institute for Artificial Intelligence(艾伦人工智能研究所)
Symbol Emergence in Cognitive Developmental Systems: a Survey
Ritsumeikan University(立命馆大学);Bogazici University(海峡大学);Czech Technical University in Prague(布拉格捷克理工大学);Queen Mary University of London(伦敦玛丽女王大学);The University of Electro-communications(电气通信大学);Council for Scientific and Industrial Research(科学与工业研究理事会);University of the Witwatersrand(金山大学);Chiba University(千叶大学);Okayama Prefectural University(冈山县立大学);Ozyegin University(厄兹耶因大学);Universität Innsbruck(因斯布鲁克大学)
Exploring the Paradigm Shift from Grounding to Skolemization for Complex Query Answering on Knowledge Graphs
School of Computer Science and Engineering, Sun Yat-sen University(中山大学计算机科学与工程学院);School of Data Science, Lingnan University(岭南大学数据科学学院);School of Science and Technology, Hong Kong Metropolitan University(香港理工大学科技学院);Department of Computing, The Hong Kong Polytechnic University(香港理工大学计算机系)
State Key Lab of Novel Software Technology, Nanjing University(南京大学计算机软件新技术国家重点实验室);University of Toronto(多伦多大学);Birkbeck, University of London(伦敦大学伯贝克学院)
Finite Groundings for ASP with Functions: A Journey through Consistency
Knowledge-Based Systems Group, TU Dresden(图腾大学达姆施塔特分校知识系统小组);LIRMM, Inria, University of Montpellier, CNRS(里尔毫米研究所、法国国家信息与自动化技术研究院、蒙彼利埃大学、国家科学研究中心);Massachusetts Institute of Technology(麻省理工学院)
Protect$^*$: Steerable Retrosynthesis through Neuro-Symbolic State Encoding
Deep Forest Sciences(深林科技);Departament de Farmacologia, Toxicologia i Química Terapèutica, Universitat de Barcelona(巴塞罗那大学药理学、毒理学与治疗化学系);Institut de Nanociència i Nanotecnologia IN2UB, Universitat de Barcelona(巴塞罗那大学纳米科学与纳米技术研究所)
Neuro-Symbolic AI for Cybersecurity: State of the Art, Challenges, and Opportunities
Department of Information Systems, University of Maryland, Baltimore County(信息系统系,马里兰大学巴尔的摩分校);Department of Computer Science and Engineering, University at Buffalo(计算机科学与工程系,布法罗大学);Department of Computer Science, University of Colorado Boulder(计算机科学系,科罗拉多大学博尔德分校);Laboratory for Cybersecurity Dynamics, Department of Computer Science, University of Colorado Colorado Springs(网络安全动力实验室,科罗拉多大学科罗拉多斯普林斯分校)
Towards Stable Symbol Grounding with Zero-Suppressed State AutoEncoder
IBM Research(IBM研究院);MIT-IBM Watson AI Lab(MIT-IBM沃森人工智能实验室)
虽然经典规划一直是人工智能的一个活跃分支,但其适用性仅限于人类精确建模的任务。完全自动化的高级代理应该能够在没有监督的情况下找到未知环境的符号表示,否则它会表现出知识获取瓶颈。与此同时,Latplan(Asai and Fukunaga 2018)通过称为状态自动编码器(SAE)的神经网络部分解决了瓶颈。SAE通过无监督学习获得基于图像的谜题域的命题表示,生成状态空间并执行经典规划。在本文中,我们确定的问题,随机行为的SAE产生的命题作为一个新的子问题的符号接地问题,符号稳定性问题。非正式地说,符号是稳定的,当它们的引用(例如命题值)不改变对小扰动的观察,和不稳定的符号是有害的符号推理。我们分析了在Latplan的问题,正式和经验,并提出了“零抑制SAE”,一个增强,稳定的命题使用封闭世界假设的想法作为前NN优化。我们发现,它发现更稳定的命题和更紧凑的表示,从而提高了Latplan的成功率。它对各种超参数具有鲁棒性,并且简化了调优工作,并且还提供了权重修剪功能作为副作用。
GRAIL: Autonomous Concept Grounding for Neuro-Symbolic Reinforcement Learning
Technical University of Darmstadt(达姆施塔特技术大学);Google Intrinsic(谷歌内在);German Research Center for Artificial Intelligence (DFKI)(德国人工智能研究中心(DFKI));Centre for Cognitive Science, TU Darmstadt(达姆施塔特技术大学认知科学中心)
Refining Manually-Designed Symbol Grounding and High-Level Planning by Policy Gradients
NEC Central Research Laboratories(NEC中央研究实验室);National Institute of Advanced Industrial Science and Technology(日本产业技术综合研究所);The University of Tokyo(东京大学)