mlsanityData to 14 Oct 2026

Concentration bounds on response-based vector embeddings of black-box generative models

arXiv 2511.08307 ↗First posted 11 Nov 2025stat.ML cs.LGcited by 1

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Generative models, such as large language models or text-to-image diffusion models, can generate relevant responses to user-given queries. Response-based vector embeddings of generative models facilitate statistical analysis and inference on a given collection of black-box generative models. The Data Kernel Perspective Space embedding is one particular method of obtaining response-based vector embeddings for a given set of generative models, already discussed in the literature. In this paper, under appropriate regularity conditions, we establish high probability concentration bounds on the sample vector embeddings for a given set of generative models, obtained through the method of Data Kernel Perspective Space embedding. Our results tell us the required number of sample responses needed in order to approximate the population-level vector embeddings with a desired level of accuracy. The algebraic tools used to establish our results can be used further for establishing concentration bounds on Classical Multidimensional Scaling embeddings in general, when the dissimilarities are observed with noise.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 13 of its 19 references that are in the map. First posted on arXiv 11 Nov 2025.

Builds on

13 of 19 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Consistent estimation of generative model representations in the data kernel perspective space25 Sep

Other references · 12

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2021Multitask Prompted Training Enables Zero-Shot Task Generalizationcited by 1,470
  4. 2022VQGAN-CLIP: Open Domain Image Generation and Editing with Natural Language Guidancecited by 285
  5. 2023Planning with Large Language Models for Code Generationcited by 182
  6. 2020Spectral Methods for Data Science: A Statistical Perspectivecited by 112
  7. 2023Perspectives on Large Language Models for Relevance Judgmentcited by 85
  8. 2016Rate-Optimal Perturbation Bounds for Singular Subspaces with Applications to High-Dimensional Statisticscited by 45
  9. 2024Evaluating the Performance of Large Language Models via Debatescited by 19
  10. 2023Comparing Foundation Models using Data Kernelscited by 14
  11. 2024Tracking the perspectives of interacting language modelscited by 14
  12. 2022Joint Spectral Clustering in Multilayer Degree-Corrected Stochastic Blockmodelscited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Query-efficient model evaluation using cached responsescited by 2

Similar, not cited

Close in wording, no citation either way

  1. 2025How Much Data Is Enough? Uniform Convergence Bounds for Generative & Vision-Language Models under Low-Dimensional Structure
  2. 2026Directional Concentration Uncertainty: A representational approach to uncertainty quantification for generative models
  3. 2025Algorithm- and Data-Dependent Generalization Bounds for Diffusion Models
  4. 2026High-Dimensional Concentration and Retrieval Instability in Embedding Spaces: Implications for Retrieval-Augmented Generation
  5. 2026Generative Embedding Benchmark: How Much Information Survives in a Dense Embedding?
  6. 2026Uniform Discrete Diffusion Models are Minimax Optimal for Estimating Distributions with Small Effective Support Size
  7. 2025On the Statistical Capacity of Deep Generative Models
  8. 2025Sample-Efficient Optimisation over the Outputs of Generative Models