Language Models Represent and Transform Concepts with Shared Geometry

• URL: http://arxiv.org/abs/2607.04525v1
• Date: Sun, 05 Jul 2026 22:16:22 GMT
• Title: Language Models Represent and Transform Concepts with Shared Geometry
• Abstract summary: How concepts are represented in neural networks is a fundamental question in machine learning.<n>We formalize concept representations as point-cloud manifold transformations and contextual transformations as vector fields.<n>Across six model families of varying scales, we find that context moves each concept differently.<n>Both the concepts being transformed and this variance structure are shared across models.
• Score: 1.5994376682356057
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: How concepts are represented in neural networks is a fundamental question in machine learning. The dominant view treats concept representations as stationary geometric objects. Yet concepts appear in context, and context transforms them. Drawing from neural population geometry, we formalize concept representations as point-cloud manifolds and contextual transformations as vector fields, and instantiate this framework in large language models. Across six model families of varying scales, we find that context moves each concept differently. The variance in these displacements is semantically organized, correlating with lexical concreteness and density. Importantly, both the concepts being transformed and this variance structure are shared across models: displacement structure transported from one model predicts held-out displacements in others significantly above chance. Together, these findings show that models share a common geometry not only in how concepts are represented, but more importantly in how context transforms them, a structure with richer organization than prior work has recognized.

Related papers

• Are You Thinking What I am Thinking? : Examining Conceptual Separation in Neural Architectures [0.0]
  We examine the conceptual organisation in Convolutional Neural Networks (CNNs) and Large Language Models (LLMs)<n>In CNNs, familiar ImageNet concepts form coherent and semantically ordered representations, while this weakens for unseen concepts.<n>In LLMs, clearly distinct domains remain well separated, related move closer together, and the distinction between ambiguous topics collapses.<n>These results suggest conceptual separation can reveal structure that output accuracy alone cannot, and may serve as a useful diagnostic of how robustly a model represents the concepts it is asked to identify.
  arXiv  Detail & Related papers  (2026-09-01T05:53:05Z)
• From Found to Designed: Concepts as a Design Axis for Large Language Models [4.979725367638729]
  We organize concept-aware interventions along two dimensions: whether concept structure is internally induced or externally grounded, and the stage of the pipeline where it is introduced.<n>This taxonomy reveals three broad patterns: inference-time approaches remain comparatively underexplored, related ideas have developed largely in isolation across pipeline stages, and externally grounded methods span the entire pipeline despite often being described under different terminology.
  arXiv  Detail & Related papers  (2026-07-29T12:18:36Z)
• HistLens: Mapping Idea Change across Concepts and Corpora [23.935379969646394]
  HistLens is a unified, SAE-based framework for multi-concept, multi-corpus conceptual-history analysis.<n>We show that HistLens supports cross-concept, cross-corpus of patterns of idea evolution and enables implicit concept computation.
  arXiv  Detail & Related papers  (2026-04-13T17:22:02Z)
• The Representational Geometry of Number [1.5994376682356057]
  We show that number representations preserve a stable relational structure across tasks.<n>We find that task-specific representations are embedded in distinct subspaces, with low-level features like magnitude encoded along separable linear directions.<n>It suggests that understanding arises when task-specific transformations are applied to a shared underlying relational structure of conceptual representations.
  arXiv  Detail & Related papers  (2026-02-06T16:35:22Z)
• Insight: Interpretable Semantic Hierarchies in Vision-Language Encoders [52.94006363830628]
  Language-aligned vision foundation models perform strongly across diverse downstream tasks.<n>Recent works decompose these representations into human-interpretable concepts, but provide poor spatial grounding and are limited to image classification tasks.<n>We propose Insight, a language-aligned concept foundation model that provides fine-grained concepts, which are human-interpretable and spatially grounded in the input image.
  arXiv  Detail & Related papers  (2026-01-20T09:57:26Z)
• A Geometric Unification of Concept Learning with Concept Cones [58.70836885177496]
  Two traditions of interpretability have evolved side by side but seldom spoken to each other: Concept Bottleneck Models (CBMs) and Sparse Autoencoders (SAEs)<n>We show that both paradigms instantiate the same geometric structure.<n>CBMs provide human-defined reference geometries, while SAEs can be evaluated by how well their learned cones approximate or contain those of CBMs.
  arXiv  Detail & Related papers  (2025-12-08T09:51:46Z)
• The Geometry of Categorical and Hierarchical Concepts in Large Language Models [15.126806053878855]
  We show how to extend the formalization of the linear representation hypothesis to represent features (e.g., is_animal) as vectors.<n>We use the formalization to prove a relationship between the hierarchical structure of concepts and the geometry of their representations.<n>We validate these theoretical results on the Gemma and LLaMA-3 large language models, estimating representations for 900+ hierarchically related concepts using data from WordNet.
  arXiv  Detail & Related papers  (2024-06-03T16:34:01Z)
• A Recursive Bateson-Inspired Model for the Generation of Semantic Formal Concepts from Spatial Sensory Data [77.34726150561087]
  This paper presents a new symbolic-only method for the generation of hierarchical concept structures from complex sensory data. The approach is based on Bateson's notion of difference as the key to the genesis of an idea or a concept. The model is able to produce fairly rich yet human-readable conceptual representations without training.
  arXiv  Detail & Related papers  (2023-07-16T15:59:13Z)
• Analyzing Encoded Concepts in Transformer Language Models [21.76062029833023]
  ConceptX analyses how latent concepts are encoded in representations learned within pre-trained language models. It uses clustering to discover the encoded concepts and explains them by aligning with a large set of human-defined concepts.
  arXiv  Detail & Related papers  (2022-06-27T13:32:10Z)
• Translational Concept Embedding for Generalized Compositional Zero-shot Learning [73.60639796305415]
  Generalized compositional zero-shot learning means to learn composed concepts of attribute-object pairs in a zero-shot fashion. This paper introduces a new approach, termed translational concept embedding, to solve these two difficulties in a unified framework.
  arXiv  Detail & Related papers  (2021-12-20T21:27:51Z)
• Image Synthesis via Semantic Composition [74.68191130898805]
  We present a novel approach to synthesize realistic images based on their semantic layouts. It hypothesizes that for objects with similar appearance, they share similar representation. Our method establishes dependencies between regions according to their appearance correlation, yielding both spatially variant and associated representations.
  arXiv  Detail & Related papers  (2021-09-15T02:26:07Z)
• Unsupervised Distillation of Syntactic Information from Contextualized Word Representations [62.230491683411536]
  We tackle the task of unsupervised disentanglement between semantics and structure in neural language representations. To this end, we automatically generate groups of sentences which are structurally similar but semantically different. We demonstrate that our transformation clusters vectors in space by structural properties, rather than by lexical semantics.
  arXiv  Detail & Related papers  (2020-10-11T15:13:18Z)

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.