Neuro-symbolic AI
- Nyhet
Foundations and Applications
Inbunden, Engelska, 2026
Av Alvaro Velasquez, Houbing Song, Pradeep Ravikumar, S. Shankar Sastry, Sandeep Neema, Alvaro (University of Colorado Boulder) Velasquez, USA) Song, Houbing (University of Maryland, MD, Pradeep (University of Texas at Austin) Ravikumar, Berkeley) Sastry, S. Shankar (University of California, S Shankar Sastry
1 869 kr
Produktinformation
- Utgivningsdatum2026-02-23
- FormatInbunden
- SpråkEngelska
- Antal sidor512
- FörlagJohn Wiley & Sons Inc
- ISBN9781394302376
Tillhör följande kategorier
- Contents1. What is Neurosymbolic AI? An Overview and Frontier Problems1.1. Introduction1.2. Neurosymbolic Artificial Intelligence1.3. Frontiers problems1.4. ConclusionBibliography 2. Reasoning in Neurosymbolic AI1.1. What is Reasoning in Neural Networks?1.2. Background: Logic and Restricted Boltzmann Machines1.3. Symbolic Reasoning with Energybased Neural Networks1.4. Logical Boltzmann Machines for MaxSAT1.5. Integrating Learning and Reasoning in Logical Boltzmann Machines1.6. Challenges for Neurosymbolic AI1.7. ConclusionBibliography 3. Neurosymbolic Assurance Using Concept Probes in Foundation Models1.1 Introduction1.2 Neural Features and Concept Probes1.3 Foundation Models as Specification Lens1.4 Symbolic Specification of ML Models Using Concept Probes1.5 Implementation and Evaluation1.6 Conclusion and Open ChallengesBibliography 4. Towards Assured Autonomy using Neurosymbolic Components and Systems1.1 Introduction1.2 Problem Formulation and Challenges: Maneuver Control for Autonomous Vehicles1.3 Software architecture: Components and Interactions1.4 Probabilistic World Model1.5 Planner1.6 Trajectory Control with Evolving Behavior Trees (EBTs)1.7 Assurance for Neuro-Symbolic Systems1.8 ConclusionsBibliography 5. Safe Neurosymbolic Learning and Control1.1. Problem Setup1.2. Hamilton-Jacobi (HJ) Reachability1.3. A NeuroSymbolic Perspective on Learning Safe Controllers1.4. Safety Assurances for Learned Controllers1.5. Frontiers, Open Questions, and Promising DirectionsBibliography 6. Controllable Generation via Locally Constrained Resampling1.1. Introduction1.2. Background1.3. Locally Constrained Resampling: A Tale of Two Distributions1.4. Related work1.5. Experimental Evaluation1.6. Conclusion and Future WorkBibliographyAppendix A: Controllable Generation via Locally Constrained Resampling 7. Tractable and Expressive Generative Modeling with Probabilistic Flow Circuits1.1. Introduction1.2. Tractable Probabilistic Modeling1.3. Probabilistic Circuits1.4. Normalizing Flows: A Primer1.5. Integrating Normalizing Flows and Probabilistic Circuits1.6. Probabilistic Flow Circuits1.7. Experiments and Results1.8. Conclusion and DiscussionAcknowledgementsBibliography 8. Toward Verifiable and Scalable In-context Fine-tuning in Neurosymbolic AI1.1 Introduction1.2 Neurosymbolic Fine-tuning Using Automated Feedback from Formal Verification1.3 Uncertainty-aware Fine-tuning and Inference for Multimodal Foundation Models1.4 Towards a Hybrid Architecture: Dynamic Interleaving of Neural and Symbolic Reasoning1.5 Conclusion and Future DirectionsBibliography 9. Physics-Informed Deep Learning1.1 IntroductionBibliography 10. Causal Representation Learning1.1. Introduction1.2. Background1.3. Interventional CRL1.4. CRL with Linear SCMs1.5. CRL with General SCMs1.6. Experiments1.7. Other approaches1.8. SummaryBibliography 11. Neuro-symbolic Computing: Hardware-Software Co-Design1.1 Introduction1.2 Background1.3 Trends and Challenges1.4 Applications and Future Topics1.5 ConclusionsBibliography 12. Programmatic Reinforcement Learning1.1. Introduction1.2. Programmatic RL1.3. Imitation-Projected Policy Gradients1.4. Related Work1.5. ConclusionBibliography 13. From Symbolic to Neuro-Symbolic Information Extraction1.1 Motivation and Overview1.2 An Example of Symbolic Information Extraction1.3 Problems of Symbolic Information Extraction Systems1.4 Generating Rules1.5 Matching Rules1.6 Take AwayBibliography 14. Neurosymbolic AI for Legal AI-TRISM: Trustworthy, Reliable, Interpretable, Safe Models1.1 Introduction1.2 Limitation of using LLM as Legal Assistant1.3 Neurosymbolic AI for Legal Domain1.4 AI-TRISM with Neurosymbolic AI1.5 Symbiosis of LLM and KG for Neurosymbolic RAG in Legal Domain1.6 Related Work1.7 AcknowledgementBibliography
Du kanske också är intresserad av
AI for Cybersecurity
Song, Houbing Song, Elisa Bertino, Alvaro Vasquez, Huihui Helen Wang, Yan Shoshitaishvili, Sumit Kumar Jha, USA) Song, Houbing (University of Maryland, MD, USA) Bertino, Elisa (Purdue University, IN, FL) Vasquez, Alvaro (University of Central Florida, USA) Wang, Huihui Helen (U.S. National Science Foundation, VA, USA) Shoshitaishvili, Yan (Arizona State University, AZ, USA) Jha, Sumit Kumar (Florida International University, FL
1 929 kr