Skip to content
Log in
Sponsored AI Networking Cookbook: Practical recipes for AI-assisted network automation and development AI Networking Cookbook: Practical recipes for AI-assisted network automation and development

Similar papers

Review Sep 2026

The Development of Compositionality in Language and Thought

Compositionality is the property of a representational system whereby complex meanings arise from the combination of simpler meaningful elements. Here we take a developmental perspective to understand the origins of compositional representational systems and their relation to language—an issue that bears on longstanding debates in cognitive science about the structure of the human mind. Based on our review, we conclude that ( a ) well-controlled evidence for compositional abilities during language acquisition remains scarce, although some studies suggest infants begin to understand specific multiword combinations during the second year of life, and ( b ) the key computation underlying compositionality—function application—can be observed in infancy during the first year of life, well before, and even outside, compositional language. These findings suggest the mind may support compositional operations prior to the emergence of compositional language, while leaving open questions about how the development of compositional thought and compositional language interact.

Unknown authors · 0 citations
Review Open access Aug 2026

Unifying the structures of language in a neural population code

It is concluded that explaining how language can emerge from neural population codes, in both biological and artificial systems, will not be achieved through the incremental refinement of algebraic-symbolic theories but will demand new theoretical paradigms.

Samuel A. Nastase, Zaid Zada, A. Goldberg et al. · 1 citation
Open access Jul 2026

Structural Composition Enables Very Fast Learning

There is a gap between neuroscientific theories of learning and the speed of learning observed in many experiments. Since the Cognitive Revolution of the 1950s, compositionality has played a central role in efforts to bridge this gap. Roughly, a compositional system is one where distinct modules are combined according to a set of rules in order to accomplish complex tasks. Recently, significant progress has been made in understanding the emergence of modules in both biological and artificial neural systems. How, and under what conditions, the rules of module recombination are represented in these systems remains an open question. Here we present a neural model that can leverage these rules to dramatically speed up learning. We first show that when faced with multiple tasks which share subcomponents, models learn a low-dimensional representation that captures how subcomponents are reused across the task set. These low-dimensional spaces encode the structure that governs how modules should be recombined. Restricting learning to these subspaces greatly reduces the amount of experience needed to acquire a novel task, even when learning from reinforcement on single trials. In some cases, we can leverage the geometric regularities of these representations to reduce learning to a form of hypothesis testing over a small set of discrete points. Finally, we use this theory to model both behavioral and neural data from non-human primates performing a compositional task, and show that key features in this data are consistent with a model in which exploration during learning is restricted to these low-dimensional spaces. Overall, this work shows that the advantages of modularity in neural systems can be greatly improved upon when models represent the structure of module reuse. Both these features working in tandem lead to learning on timescales similar to biological intelligences, and hence provide a model for how such fast, adaptable behavior can emerge from systems of neurons.

Reidar Riveland, Alexandre Pouget, P. Latham · 0 citations
Preprint Aug 2026

Mathematical Principles and Experimental Discoveries of the Emergence of Symbolic Patterns in Artificial Neural Networks

Artificial Neural networks (ANNs) are often treated as black-box models, making explainability a central challenge in deep learning. Many engineering methods have been proposed to approximately explain the ANN from various perspectives, such as feature attribution and visualization. However, it remains a long-standing open question whether the complex inference logic of an ANN can be explained exhaustively and concisely as sparse symbolic patterns. This raises a deeper inquiry: does the emergence of symbolic patterns reflect a natural law rather than chance? Here, we show that across a broad class of ANNs trained on diverse tasks, their inference logic can indeed be reformulated as sparse symbolic interactions. We further prove that two common mathematical criteria, which are implicitly required across tasks, lead to the emergence of such sparse symbolic interactions. Empirical evidence confirms that the two criteria hold for the majority of input samples in diverse models. Furthermore, the faithfulness of these interactions is also demonstrated by their strong sample-to-sample and model-to-model transferability, as well as their ability to explain the overall generalization power of ANNs. Our theoretical analysis and extensive experiments provide a solid foundation for symbolic explanations of ANNs, and offer novel insights into the ANN's generalization power. Our findings also highlight the potential of communicative learning, a paradigm in which the inference logic of an ANN can be directly inspected and tuned at the level of symbolic patterns, thus complementing traditional end-to-end learning paradigm. Finally, the observed emergence of symbolic patterns in ANNs suggests that similar symbolic representations may also emerge in other types of black-box systems under certain conditions, because our proof does not depend on any specific ANN architecture.

Quanshi Zhang, Qihan Ren, Siyu Lou · 0 citations
Aug 2026

Necessity and plurality of neural dynamics under ROSE.

I am grateful to the authors of the commentary articles for identifying productive points of pressure for any neurocomputational account of syntax: whether semantic interpretation can proceed without a full syntactic derivation; how dynamical motifs are selected and coordinated; whether proposed mechanisms generalize across languages and modalities; how evolution may have repurposed older memory circuitry; and how neural dynamics distinguish types and tokens. Here, I clarify that ROSE is an architecture for implementing hierarchical syntactic computation, bringing with it no commitment that its full code is obligatory for every act of meaning construction. The commentaries suggest a number of compelling concrete extensions: task-dependent gating of S/E; factorized R/O subspaces; representation-relative rather than language-specific complexity measures; comparative tests of neural reuse; and occurrence-sensitive phase or state-space addresses. These thoughtful extensions sharpen the central aim of ROSE to formulate multiscale, falsifiable links between formal properties of language and neural dynamics.

Elliot Murphy · 0 citations
Open access Jul 2026

Processing Negation May Demonstrate Compositionality in Rhesus Macaques

Compositionality, the ability to flexibly combine meaningful words into higher-level representations, is a critical feature of human language. However, the evolutionary origins of compositional processing are unclear, and evidence for this ability in nonhuman primates is largely lacking. Compositionality is inherent in processing negative clauses, such as ‘not red’. Here, we conducted a series of three experiments using a novel, nonlinguistic ‘negation’ paradigm, which provide evidence of compositional processing in nonhuman primates. Macaques learned to associate an abstract visual stimulus with the process of negation (i.e., ‘not’), and to combine this compositionally with novel images (e.g., a dog) on a trial-by-trial basis (i.e., to select a stimulus that was not a dog). Our results suggest compositional processing in monkeys, raising the possibility that the core of this language-critical ability may be present in monkeys, and thus may not be unique to humans or to language.

Angelle Antoun, Rohini Murugan, Kathleen J. Bostick et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.