mlsanityData to 14 Oct 2026

GRACE: LLM-Grounded Semantic Metric Spaces for Scalable Mixed-Data Clustering

arXiv 2608.07881 ↗First posted 8 Aug 2026cs.AI cs.CL cs.ITcited by 0

Not on the map. Why?

Abstract (arXiv)

Clustering mixed tabular data requires a unified metric space to bridge the inherent heterogeneity between continuous numerical measurements and discrete categorical symbols. Traditionally, algorithms rely entirely on dataset-internal statistics to estimate categorical relationships, which confines the learned metric to empirical co-occurrences and ignores conceptually obvious yet statistically unobserved affinities. Although LLMs offer external world knowledge, applying their text-centric reasoning to highly abstract tabular concepts presents significant challenges. Bridging this modality gap to construct a semantically complete metric typically requires embedding LLMs into iterative metric learning loops to dynamically optimize cross-modality representations. This incurs intractable computational overhead, forcing a compromise between semantic enrichment and scalability. Therefore, we propose GRACE, an LLM-grounded framework for scalable mixed-data clustering. GRACE shifts semantic acquisition to the attribute-value level via a multi-perspective LLM querying strategy, mapping heterogeneous values into knowledge-informed descriptions. Crucially, this one-shot grounding extracts general-purpose semantic representations that embed heterogeneous attributes into a unified space, decoupling expensive LLM invocation from iterative optimization. Furthermore, GRACE cross-validates these external semantics against dataset-internal statistical evidence to ensure alignment with the dataset-specific cluster structure. Ultimately, GRACE matches the scalability of conventional statistics-driven baselines while achieving superior clustering accuracy and conceptual interpretability over 11 competing methods. The source code is available at https://github.com/develop-yang/GRACE-GRACE-A

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 28 of its 52 references that are in the map. First posted on arXiv 8 Aug 2026.

Builds on

28 of 52 references.

References · 28

  1. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  2. 2021SimCSE: Simple Contrastive Learning of Sentence Embeddingscited by 2,173
  3. 2021Revisiting Deep Learning Models for Tabular Datacited by 740
  4. 2022Language Models are Realistic Tabular Data Generatorscited by 294
  5. 2022TabLLM: Few-shot Classification of Tabular Data with Large Language Modelscited by 277
  6. 2023Large Language Models for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineeringcited by 127
  7. 2021Supporting Clustering with Contrastive Learningcited by 90
  8. 2016Entity Embeddings of Categorical Variablescited by 87
  9. 2024Large Language Models Can Automatically Engineer Features for Few-Shot Tabular Learningcited by 81
  10. 2023Large Language Models Enable Few-Shot Clusteringcited by 72
  11. 2023ClusterLLM: Large Language Models as a Guide for Text Clusteringcited by 70
  12. 2024TabDiff: a Mixed-type Diffusion Model for Tabular Data Generationcited by 63
16 more
  1. 2023Image Clustering with External Guidancecited by 36
  2. 2023Interpretable Deep Clustering for Tabular Datacited by 19
  3. 2024Hierarchical Vector Quantization for Unsupervised Action Segmentationcited by 19
  4. 2023Generalized Category Discovery with Large Language Models in the Loopcited by 15
  5. 2020Unsupervised Heterogeneous Coupling Learning for Categorical Representationcited by 13
  6. 2021A Simple Approach to Automated Spectral Clusteringcited by 10
  7. 2026Robust Categorical Data Clustering Guided by Multi-Granular Competitive Learningcited by 7
  8. 2024Text-Guided Image Clusteringcited by 5
  9. 2024Categorical Data Clustering via Value Order Estimated Distance Metric Learningcited by 5
  10. 2025Break the Tie: Learning Cluster-Customized Category Relationships for Categorical Data Clusteringcited by 4
  11. 2026Learning Unified Distance Metric for Heterogeneous Attribute Data Clusteringcited by 4
  12. 2026Learning Order Forest for Qualitative-Attribute Data Clusteringcited by 3
  13. 2024Information-Theoretic Generative Clustering of Documentscited by 2
  14. 2026Bridging the Semantic Gap for Categorical Data Clustering via Large Language Modelscited by 2
  15. 2025SpeHeatal: A Cluster-Enhanced Segmentation Method for Sperm Morphology Analysiscited by 1
  16. 2025Mixed Data Clustering Survey and Challengescited by 1

Built on by

Not cited yet.