16 – 18 September 2026 · Universitat Politècnica de València
Provisional programme. Times and the fourth keynote are subject to change; the final version will be published before the conference.
Welcome and Opening
Prof. Stephen Muggleton FREng (Imperial College London and Nanjing University)
Are AI's key challenges solved?
Generative Pre-trained Transformer models (GPT) are a form of Large-Language-Model (LLM) which has attracted wide-scale interest based on general open-ended query-answering. In this talk I will look at remaining challenges for the field of Artificial Intelligence. In particular, I will argue that although widely available LLM systems partially address Alan Turing's “Imitation Game” [1950 paper in the Journal Mind], now called the “Turing Test”, they fall short of the “Super Criticality Challenge” which Turing provided within the same paper. While Turing's paper initiated modern discussions on these topics, it was John McCarthy [1956] who named the field “Artificial Intelligence”. In a 2006 keynote speech at the Inductive Logic Programming conference (Santiago de Compostela), McCarthy introduced a “Discovery Challenge” which is closely related to Turing's Supercriticality. I will exemplify a range of human discoveries in Science, Engineering and Mathematics, and argue that neither Turing's nor McCarthy's challenges are addressed by existing LLM techniques. By contrast, to enable progress on Discovery Systems, further work is required on development of methods for identifying and explaining rare phenomena. Some initial work and directions for further studies will be described.
#8Ultra Strong Machine Learning: LLM-Generated Explanations Do Not Yet Suffice for Teaching Humans Active Learning Strategy
EMBL-EBI; University of Bamberg; Imperial College London
Coffee break
#1Explainable Symbolic Constraint Reasoner
LIPN, Université Sorbonne Paris Nord; NukkAI
#9Learning Symbolic Constraint Representations from Examples: A Neuro-Symbolic Approach
SIMULA; LISN CNRS, Paris-Saclay University
#13Assessing Reliability of Symbol Detection in Concept Bottleneck Models
Public University of Navarre; University of Essex
#20Enhancing Explainable AI: A Gaussian Mixture Approach to Numerical-Symbolic Reasoning
University of Surrey
#18An Investigation into Interpretable Embeddings Resulting from Large Logical Rule Sets
Institute for Application-Oriented Knowledge Processing (FAW), JKU
Lunch
Laurent Orseau (Google DeepMind)
Levin Tree Search: Search-and-Learn with Formal Guarantees via Time Sharing
Solving complex combinatorial problems from scratch requires a tight integration between learning and search. This keynote presents the Levin Tree Search (LTS) family of algorithms and analyzes their performance guarantees using the simple yet fundamental concept of time-sharing. We show how allocating computational budgets based on policy and heuristic quality can be used to provide strict bounds on the number of node visits before finding a solution. We then describe how these algorithms operate within a search-and-learn bootstrap paradigm to solve combinatorial problems without prior knowledge, using only a training set of unsolved instances of varied difficulty. We introduce in particular the √LTS (root-LTS) algorithm, which extends the time-sharing perspective to all nodes in a search tree. By distributing the computational budget among multiple concurrent searches rooted at different nodes, √LTS implicitly decomposes the original problem into smaller subtasks. We present formal guarantees showing that this rerooting mechanism provides exponential speedups over standard LTS. Speaker bio
#27Bottom-Clause Reduction Using Definitional Redundancy for Efficient Hypothesis Search
University of Surrey
#7FedPopper: Federated Learning Logic Programs from Aggregated Failures
UNamur
Orxata
Guided visit
Kevin Ellis (Cornell University)
Abstract World Models for Learning and Reasoning
Consider experimenting to learn how to use a new appliance, webpage, or toy: Within tens of minutes, we can learn how something new works, and use that knowledge to achieve novel goals, make sense of similar devices, and communicate our new knowledge in natural language. How could an AI system similarly acquire, transfer, and communicate its knowledge of how things work? To make progress on this question, we study AI systems which learn abstract world knowledge represented as symbolic programs written in Python, natural language, and/or logic. To test these systems, we assemble new interactive benchmarks focusing on interactive puzzle-like grid worlds, and simulated robotic environments, where the agent must interact with a new environment, and then learn enough about how it works to answer questions about the dynamics or achieve novel goals. Speaker bio
#26Teaching a Minimalist Machine to Discover Recursive Programs for Arithmetic
Centre for Cognitive Science & Institute of Psychology, Technische Universität Darmstadt; Honda Research Institute Europe GmbH
#24Learning Probabilistic Logic Programs with Functional Gradient Guided Language Models
Technische Universität Darmstadt; The University of Texas at Dallas; Oregon State University; AT & T
Coffee break
#4ALBI: LLM-Guided Inductive Logic Programming via Iterative Generate-Test-Constrain Refinement
Tallinn University of Technology
#16From Narratives to Reasoning: Can LLMs Generate Reasoning-Ready Logic Forms?
Miami University
#23Kinship Reasoning as a Testbed for Neuro-Symbolic Architectures: Comparing LLM–Prolog Integration Strategies and Assessing Cultural Bias
University of York
#11From Words to Actions: Grounding LLM Plans in Real-Time Environments via Neuro-Symbolic Logic Programming
Nanjing University
#31Bridging LLM-based Argumentation and Formal Reasoning via Categorized Attack Extraction and AAF Semantics
Johannes Gutenberg University Mainz
Lunch
#19Neuro-Symbolic Reasoning for Social Deduction Games: A Case Study on Blood on the Clocktower
Nanjing University
#6Query-Driven Learning for Lifted Reasoning
Washington University in St Louis; Washington University in St. Louis
#29ILP Meets RDF: A Data Transformation Framework for Popper and AMIE Interoperability
Prague University of Economics and Business
#30Fuzzy First-Order Logical Soft Decision Trees
Johannes Gutenberg University
#39WAMpy: Efficient Synthesis of Prolog Programs in PythonDemo
Centre for Cognitive Science & Institute of Psychology, Technische Universität Darmstadt
#40Vocabulary-Agnostic Ranking for ILP Candidate ExecutionLate-breaking Paper
University of Bristol
#33Predicate Renaming via Large Language ModelsRecently Published Paper
University of Ferara; Nantes Université; University of Ferrara; National Institute of Informatics
Social dinner
Jennifer Neville (Microsoft Research, Purdue University)
The Complexity Gap: Understanding AI Behavior in Realistic Settings
Modern AI systems achieve impressive performance on benchmarks composed of isolated, fully specified tasks. Real-world knowledge work, however, is rarely so clean. Users communicate goals through evolving conversations, leave important details implicit, revisit prior decisions, and combine tasks of varying complexity into long-horizon workflows. As complexity accumulates through under-specification, changing intent, long-range dependencies, and iterative transformations of content, AI systems can exhibit subtle failures that are difficult for both users and developers to anticipate, diagnose, and correct. In this talk, I will discuss recent efforts to move beyond traditional benchmarks and study AI behavior in realistic knowledge-work settings. Drawing on work spanning complex reasoning tasks, multi-turn conversations, and long-horizon workflows, I will present empirical and theoretical results that reveal recurring failure modes as complexity increases. Together, these findings provide a foundation for understanding how and why AI behavior changes in realistic settings, and for developing evaluation paradigms that better reflect the challenges of real-world knowledge work. Speaker bio
#17Control-by-Structure: Certificate-Gated Agent Adaptation under Task Transformations
LITIS UR 4108- INSA de Rouen - Normandie University
#25Neurosymbolic Imitation Learning with Human Guidance: A Privileged Information Approach
The University of Texas At Dallas; TU Darmstadt
Coffee break
#14Utility-Driven Multi-Mapping Transfer Learning for Relational Domains
UFRJ; University of Avignon; UERJ
#15Choose your Graph: Dataset Selection for GNN and SRL Transfer Learning for Graph Data
Universidade Federal do Rio de Janeiro; Universidade Federal Fluminense
#10HyDI: A hybrid Deep Learning-Inductive Logic Programming ensemble for multi-label classification
University of Osnabrück; Osnabrück University
#28Preserving Biological Semantics: Relational Rule Learning for Media Prediction under Incomplete Evidence
Prague University of Economics and Business; Lawrence Berkeley National Laboratory
#22Deep probabilistic logic programming for diagnostic reasoning from incomplete information: A case study in stroke detection
German University of Digital Science
Lunch
#34Attention Cannot Orient: A Symmetric Kernel Obstruction and Its Lorentzian Repair in Neuro-Symbolic GroundingLate-breaking Paper
University of Bucharest
#36Inductive Reasoning for Weakly Unsupervised Symbol Grounding from ImagesLate-breaking Paper
Zhongguancun Academy; Vienna University of Technology (TU Wien); National Institute of Informatics
#37DIFFERENTIABLE RULE INDUCTION FROM RAW SEQUENCE INPUTSRecently Published Paper
Zhongguancun Academy; National Institute of Informatics
#38Adaptive Data-Knowledge Alignment in Genetic Perturbation PredictionRecently Published Paper
KU Leuven; EMBL-EBI
#41The Use of Retinal Vascular Features for Disease Diagnosis using Inductive logic programmingLate-breaking Paper
University of Surrey
#42Phase-Aware Interpretable Representations for Cyclical Time SeriesLate-breaking Paper
University of Bristol
Closing session and community meeting
Full papers are allocated 20 minutes including questions. Short papers — late-breaking papers, recently published papers and demos — are allocated 15 minutes. Session chairs will be announced closer to the conference.