The Mechanistic Emergence of Symbol Grounding in Language Models

• URL: http://arxiv.org/abs/2510.13796v2
• Date: Thu, 16 Oct 2025 17:51:48 GMT
• Title: The Mechanistic Emergence of Symbol Grounding in Language Models
• Authors: Shuyu Wu, Ziqiao Ma, Xiaoxi Luo, Yidong Huang, Josue Torres-Fonseca, Freda Shi, Joyce Chai,
• Abstract summary: Symbol grounding describes how symbols acquire their meanings by connecting to real-world sensorimotor experiences.<n>Recent work has shown preliminary evidence that grounding may emerge in (vision-language) models trained at scale without using explicit grounding objectives.<n>Our results provide behavioral and mechanistic evidence that symbol grounding can emerge in language models.
• Score: 27.407379293112587
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Symbol grounding (Harnad, 1990) describes how symbols such as words acquire their meanings by connecting to real-world sensorimotor experiences. Recent work has shown preliminary evidence that grounding may emerge in (vision-)language models trained at scale without using explicit grounding objectives. Yet, the specific loci of this emergence and the mechanisms that drive it remain largely unexplored. To address this problem, we introduce a controlled evaluation framework that systematically traces how symbol grounding arises within the internal computations through mechanistic and causal analysis. Our findings show that grounding concentrates in middle-layer computations and is implemented through the aggregate mechanism, where attention heads aggregate the environmental ground to support the prediction of linguistic forms. This phenomenon replicates in multimodal dialogue and across architectures (Transformers and state-space models), but not in unidirectional LSTMs. Our results provide behavioral and mechanistic evidence that symbol grounding can emerge in language models, with practical implications for predicting and potentially controlling the reliability of generation.

Related papers

• Emergent Structured Representations Support Flexible In-Context Inference in Large Language Models [77.98801218316505]
  Large language models (LLMs) exhibit emergent behaviors suggestive of human-like reasoning.<n>We investigate the internal processing of LLMs during in-context concept inference.
  arXiv  Detail & Related papers  (2026-02-08T03:14:39Z)
• Stable Language Guidance for Vision-Language-Action Models [62.80963701282789]
  Residual Semantic Steering is a probabilistic framework that disentangles physical affordance from semantic execution.<n> RSS achieves state-of-the-art robustness, maintaining performance even under adversarial linguistic perturbations.
  arXiv  Detail & Related papers  (2026-01-07T16:16:10Z)
• On measuring grounding and generalizing grounding problems [0.0]
  The symbol grounding problem asks how tokens cat can be about cats, as opposed to mere shapes manipulated in calculus.<n>We recast grounding from a binary judgment into an audit across desiderata, each indexed by an evaluation.<n>We apply this framework to four grounding modes (symbolic; referential; vectorial; relational) and three case studies.
  arXiv  Detail & Related papers  (2025-12-05T22:58:47Z)
• GeoGNN: Quantifying and Mitigating Semantic Drift in Text-Attributed Graphs [59.61242815508687]
  Graph neural networks (GNNs) on text--attributed graphs (TAGs) encode node texts using pretrained language models (PLMs) and propagate these embeddings through linear neighborhood aggregation.<n>This work introduces a local PCA-based metric that measures the degree of semantic drift and provides the first quantitative framework to analyze how different aggregation mechanisms affect manifold structure.
  arXiv  Detail & Related papers  (2025-11-12T06:48:43Z)
• Less is More: Local Intrinsic Dimensions of Contextual Language Models [15.934492316122025]
  We introduce a novel perspective based on the geometric properties of contextual latent embeddings to study the effects of training and fine-tuning.<n>We show that the local dimensions provide insights into the model's training dynamics and generalization ability.<n>Our experiments suggest configuring a practical: reductions in the mean local dimension tend to accompany and predict subsequent performance gains.
  arXiv  Detail & Related papers  (2025-06-01T14:30:46Z)
• Mechanistic Understanding and Mitigation of Language Confusion in English-Centric Large Language Models [56.61984030508691]
  We present the first mechanistic interpretability study of language confusion.<n>We show that confusion points (CPs) are central to this phenomenon.<n>We show that editing a small set of critical neurons, identified via comparative analysis with a multilingual-tuned counterpart, substantially mitigates confusion.
  arXiv  Detail & Related papers  (2025-05-22T11:29:17Z)
• Reasoning Circuits in Language Models: A Mechanistic Interpretation of Syllogistic Inference [13.59675117792588]
  Recent studies on language models (LMs) have sparked a debate on whether they can learn systematic inferential principles.<n>This paper presents a mechanistic interpretation of syllogistic inference.
  arXiv  Detail & Related papers  (2024-08-16T07:47:39Z)
• Language, Environment, and Robotic Navigation [0.0]
  We propose a unified framework where language functions as an abstract communicative system and as a grounded representation of perceptual experiences. Our review of cognitive models of distributional semantics and their application to autonomous agents underscores the transformative potential of language-integrated systems.
  arXiv  Detail & Related papers  (2024-04-03T20:30:38Z)
• Discrete, compositional, and symbolic representations through attractor dynamics [51.20712945239422]
  We introduce a novel neural systems model that integrates attractor dynamics with symbolic representations to model cognitive processes akin to the probabilistic language of thought (PLoT) Our model segments the continuous representational space into discrete basins, with attractor states corresponding to symbolic sequences, that reflect the semanticity and compositionality characteristic of symbolic systems through unsupervised learning, rather than relying on pre-defined primitives. This approach establishes a unified framework that integrates both symbolic and sub-symbolic processing through neural dynamics, a neuroplausible substrate with proven expressivity in AI, offering a more comprehensive model that mirrors the complex duality of cognitive operations
  arXiv  Detail & Related papers  (2023-10-03T05:40:56Z)
• From Word Models to World Models: Translating from Natural Language to the Probabilistic Language of Thought [124.40905824051079]
  We propose rational meaning construction, a computational framework for language-informed thinking. We frame linguistic meaning as a context-sensitive mapping from natural language into a probabilistic language of thought. We show that LLMs can generate context-sensitive translations that capture pragmatically-appropriate linguistic meanings. We extend our framework to integrate cognitively-motivated symbolic modules.
  arXiv  Detail & Related papers  (2023-06-22T05:14:00Z)
• Models of symbol emergence in communication: a conceptual review and a guide for avoiding local minima [0.0]
  Computational simulations are a popular method for testing hypotheses about the emergence of communication. We identify the assumptions and explanatory targets of several most representative models and summarise the known results. In line with this perspective, we sketch the road towards modelling the emergence of meaningful symbolic communication.
  arXiv  Detail & Related papers  (2023-03-08T12:53:03Z)
• Color Overmodification Emerges from Data-Driven Learning and Pragmatic Reasoning [53.088796874029974]
  We show that speakers' referential expressions depart from communicative ideals in ways that help illuminate the nature of pragmatic language use. By adopting neural networks as learning agents, we show that overmodification is more likely with environmental features that are infrequent or salient.
  arXiv  Detail & Related papers  (2022-05-18T18:42:43Z)
• Emergent Communication with World Models [80.55287578801008]
  We introduce Language World Models, a class of language-conditional generative model which interpret natural language messages. We incorporate this "observation" into a persistent memory state, and allow the listening agent's policy to condition on it. We show this improves effective communication and task success in 2D gridworld speaker-listener navigation tasks.
  arXiv  Detail & Related papers  (2020-02-22T02:34:51Z)

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.