Grounding vs. Compositionality: On the Non-Complementarity of Reasoning in Neuro-Symbolic Systems

• URL: http://arxiv.org/abs/2604.26521v1
• Date: Wed, 29 Apr 2026 10:41:30 GMT
• Title: Grounding vs. Compositionality: On the Non-Complementarity of Reasoning in Neuro-Symbolic Systems
• Abstract summary: A central assumption in neuro-symbolic AI is that compositional reasoning will emerge as a byproduct of successful symbol grounding.<n>This work presents the first systematic empirical analysis to challenge this assumption by disentangling the contributions of grounding and reasoning.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Compositional generalization remains a foundational weakness of modern neural networks, limiting their robustness and applicability in domains requiring out-of-distribution reasoning. A central, yet unverified, assumption in neuro-symbolic AI is that compositional reasoning will emerge as a byproduct of successful symbol grounding. This work presents the first systematic empirical analysis to challenge this assumption by disentangling the contributions of grounding and reasoning. To operationalize this investigation, we introduce the Iterative Logic Tensor Network ($i$LTN), a fully differentiable architecture designed for multi-step deduction. Using a formal taxonomy of generalization -- probing for novel entities, unseen relations, and complex rule compositions -- we demonstrate that a model trained solely on a grounding objective fails to generalize. In contrast, our full $i$LTN, trained jointly on perceptual grounding and multi-step reasoning, achieves high zero-shot accuracy across all tasks. Our findings provide conclusive evidence that symbol grounding, while necessary, is insufficient for generalization, establishing that reasoning is not an emergent property but a distinct capability that requires an explicit learning objective.

Related papers

• NeurOWL: An LLM-Based Neural-symbolic Framework for Incomplete OWL Ontology Reasoning [15.085939995703667]
  We focus on a fundamental subsumption reasoning problem.<n>We propose NeurOWL, an end-to-end neurosymbolic framework.<n>We evaluate NeurOWL on realworld- NeurOWL across multiple domains.
  arXiv  Detail & Related papers  (2026-07-17T09:12:10Z)
• Weakly Supervised Concept Learning for Object-centric Visual Reasoning [3.6959117433332724]
  Neurosymbolic systems promise to combine deep neural network's (DNN) processing of raw sensor inputs with few-shot performance of symbolic artificial intelligence.<n>This paper introduces an efficient weak supervision scheme for the perception stage to ground its output symbols for logical induction in object-centric reasoning tasks.
  arXiv  Detail & Related papers  (2026-05-05T22:50:43Z)
• Hybrid Models for Natural Language Reasoning: The Case of Syllogistic Logic [3.421904493396495]
  We investigate the logical generalization capabilities of pre-trained large language models (LLMs) using the syllogistic fragment as a benchmark.<n>We propose a hybrid architecture integrating symbolic reasoning with neural computation.<n>Our experiments show that high efficiency is preserved even with relatively small neural components.
  arXiv  Detail & Related papers  (2025-10-10T15:27:29Z)
• Propositional Logic for Probing Generalization in Neural Networks [3.6037930269014633]
  We investigate the generalization behavior of three key neural architectures (Transformers, Graph Convolution Networks and LSTMs) in a controlled task rooted in propositional logic.<n>We find thatTransformers fail to apply negation compositionally, unless structural biases are introduced.<n>Our findings highlight persistent limitations in the ability of standard architectures to learn systematic representations of logical operators.
  arXiv  Detail & Related papers  (2025-06-10T16:46:05Z)
• Failure Modes of LLMs for Causal Reasoning on Narratives [51.19592551510628]
  We investigate the interaction between world knowledge and logical reasoning.<n>We find that state-of-the-art large language models (LLMs) often rely on superficial generalizations.<n>We show that simple reformulations of the task can elicit more robust reasoning behavior.
  arXiv  Detail & Related papers  (2024-10-31T12:48:58Z)
• Feature contamination: Neural networks learn uncorrelated features and fail to generalize [5.642322814965062]
  Learning representations that generalize under distribution shifts is critical for building robust machine learning models.<n>We show that even allowing a neural network to explicitly fit the representations obtained from a teacher network that can generalize out-of-distribution is insufficient for the generalization of the student network.
  arXiv  Detail & Related papers  (2024-06-05T15:04:27Z)
• Semantic Loss Functions for Neuro-Symbolic Structured Prediction [74.18322585177832]
  We discuss the semantic loss, which injects knowledge about such structure, defined symbolically, into training. It is agnostic to the arrangement of the symbols, and depends only on the semantics expressed thereby. It can be combined with both discriminative and generative neural models.
  arXiv  Detail & Related papers  (2024-05-12T22:18:25Z)
• LOGICSEG: Parsing Visual Semantics with Neural Logic Learning and Reasoning [73.98142349171552]
  LOGICSEG is a holistic visual semantic that integrates neural inductive learning and logic reasoning with both rich data and symbolic knowledge. During fuzzy logic-based continuous relaxation, logical formulae are grounded onto data and neural computational graphs, hence enabling logic-induced network training. These designs together make LOGICSEG a general and compact neural-logic machine that is readily integrated into existing segmentation models.
  arXiv  Detail & Related papers  (2023-09-24T05:43:19Z)
• Logical Neural Networks [51.46602187496816]
  We propose a novel framework seamlessly providing key properties of both neural nets (learning) and symbolic logic (knowledge and reasoning) Every neuron has a meaning as a component of a formula in a weighted real-valued logic, yielding a highly intepretable disentangled representation. Inference is omni rather than focused on predefined target variables, and corresponds to logical reasoning.
  arXiv  Detail & Related papers  (2020-06-23T16:55:45Z)
• Compositional Generalization by Learning Analytical Expressions [87.15737632096378]
  A memory-augmented neural model is connected with analytical expressions to achieve compositional generalization. Experiments on the well-known benchmark SCAN demonstrate that our model seizes a great ability of compositional generalization.
  arXiv  Detail & Related papers  (2020-06-18T15:50:57Z)
• Relational reasoning and generalization using non-symbolic neural networks [66.07793171648161]
  Previous work suggested that neural networks were not suitable models of human relational reasoning because they could not represent mathematically identity, the most basic form of equality. We find neural networks are able to learn basic equality (mathematical identity), (2) sequential equality problems (learning ABA-patterned sequences) with only positive training instances, and (3) a complex, hierarchical equality problem with only basic equality training instances. These results suggest that essential aspects of symbolic reasoning can emerge from data-driven, non-symbolic learning processes.
  arXiv  Detail & Related papers  (2020-06-14T18:25:42Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.