# Anatomy of a Machine Learning Ecosystem: 2 Million Models on Hugging Face

arXiv: https://arxiv.org/abs/2508.06811 · HTML: https://mlsanity.com/paper/2508.06811/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 54 references that are in the map.
- First posted on arXiv 9 Aug 2025.

## Builds on (most cited first, 25 of 25)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019)
- [node2vec: Scalable Feature Learning for Networks](https://mlsanity.com/paper/1607.00653/) (2016)
- [Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models](https://mlsanity.com/paper/2506.05176/) (2025)
- [Model Cards for Model Reporting](https://mlsanity.com/paper/1810.03993/) (2018)
- [The Flan Collection: Designing Data and Methods for Effective Instruction Tuning](https://mlsanity.com/paper/2301.13688/) (2023)
- [Representation Learning on Graphs: Methods and Applications](https://mlsanity.com/paper/1709.05584/) (2017)
- [Text-to-image Diffusion Models in Generative AI: A Survey](https://mlsanity.com/paper/2303.07909/) (2023)
- [A Phase Transition in Diffusion Models Reveals the Hierarchical Nature of Data](https://mlsanity.com/paper/2402.16991/) (2024)
- [The Data Provenance Initiative: A Large Scale Audit of Dataset Licensing & Attribution in AI](https://mlsanity.com/paper/2310.16787/) (2023)
- [Causal Reasoning from Meta-reinforcement Learning](https://mlsanity.com/paper/1901.08162/) (2019)
- [A Survey on Contextual Embeddings](https://mlsanity.com/paper/2003.07278/) (2020)
- [Dislocated Accountabilities in the AI Supply Chain: Modularity and Developers' Notions of Responsibility](https://mlsanity.com/paper/2209.09780/) (2022)
- [Ecosystem Graphs: The Social Footprint of Foundation Models](https://mlsanity.com/paper/2303.15772/) (2023)
- [Fine-Tuning Games: Bargaining and Adaptation for General-Purpose Models](https://mlsanity.com/paper/2308.04399/) (2023)
- [Accounting for AI and Users Shaping One Another: The Role of Mathematical Models](https://mlsanity.com/paper/2404.12366/) (2024)
- [HuggingGraph: Understanding the Supply Chain of LLM Ecosystem](https://mlsanity.com/paper/2507.14240/) (2025)
- [AI Supply Chains: An Emerging Ecosystem of AI Actors, Products, and Services](https://mlsanity.com/paper/2504.20185/) (2025)
- [Selective Response Strategies for GenAI](https://mlsanity.com/paper/2502.00729/) (2025)
- [Low-resource Neural Machine Translation with Cross-modal Alignment](https://mlsanity.com/paper/2210.06716/) (2022)
- [Safety vs. Performance: How Multi-Objective Learning Reduces Barriers to Market Entry](https://mlsanity.com/paper/2409.03734/) (2024)
- [The Backfiring Effect of Weak AI Safety Regulation](https://mlsanity.com/paper/2503.20848/) (2025)
- [Speeding up Model Loading with fastsafetensors](https://mlsanity.com/paper/2505.23072/) (2025)
- [Modeling the Economic Impacts of AI Openness Regulation](https://mlsanity.com/paper/2507.14193/) (2025)

## Built on by (most cited first, 12 of 12)

- [Efficient Evaluation of LLM Performance with Statistical Guarantees](https://mlsanity.com/paper/2601.20251/) (2026)
- [Economies of Open Intelligence: Tracing Power & Participation in the Model Ecosystem](https://mlsanity.com/paper/2512.03073/) (2025)
- [EVOMAL: Self-Poisoning in Self-Evolving Coding Agents](https://mlsanity.com/paper/2608.25776/) (2026)
- [A Cartography of Open Collaboration in Open Source AI: Mapping Practices, Motivations, and Governance in 14 Open Large Language Model Projects](https://mlsanity.com/paper/2509.25397/) (2025)
- [Which Models Are Our Models Built On? Auditing Invisible Dependencies in Modern LLMs](https://mlsanity.com/paper/2606.12385/) (2026)
- [AdaQE-CG: Adaptive Query Expansion for Web-Scale Generative AI Model and Data Card Generation](https://mlsanity.com/paper/2604.09617/) (2026)
- [TensorDex: A Compact, Tensor-Centric Storage System for Modern AI Models](https://mlsanity.com/paper/2604.17104/) (2026)
- [Cost-Ordered Feasibility for Multi-Armed Bandits with Cost Subsidy](https://mlsanity.com/paper/2605.07171/) (2026)
- [ArtifactLinker: Linking Scientific Artifacts for Automatic State-of-the-Art Discovery](https://mlsanity.com/paper/2605.16902/) (2026)
- [A governance horizon for ethical-use constraints in open-weight AI models](https://mlsanity.com/paper/2605.24383/) (2026)
- [AI Supply Chain Galaxy: 3D Visual Analytics for License Compliance](https://mlsanity.com/paper/2606.16292/) (2026)
- [The evolution of sex for artificial intelligence: a population-genetic framework for multigenerational model populations](https://mlsanity.com/paper/2609.18560/) (2026)

## Abstract (arXiv)

Many have observed that the development and deployment of generative machine learning (ML) and artificial intelligence (AI) models follow a distinctive pattern in which pre-trained models are adapted and fine-tuned for specific downstream tasks. However, there is limited empirical work that examines the structure of these interactions. This paper analyzes 1.86 million models on Hugging Face, a leading peer production platform for model development. Our study of model family trees -- networks that connect fine-tuned models to their base or parent -- reveals sprawling fine-tuning lineages that vary widely in size and structure. Using an evolutionary biology lens to study ML models, we use model metadata and model cards to measure the genetic similarity and mutation of traits over model families. We find that models tend to exhibit a family resemblance, meaning their genetic markers and traits exhibit more overlap when they belong to the same model family. However, these similarities depart in certain ways from standard models of asexual reproduction, because mutations are fast and directed, such that two `sibling' models tend to exhibit more similarity than parent/child pairs. Further analysis of the directional drifts of these mutations reveals qualitative insights about the open machine learning ecosystem: Licenses counter-intuitively drift from restrictive, commercial licenses towards permissive or copyleft licenses, often in violation of upstream license's terms; models evolve from multi-lingual compatibility towards english-only compatibility; and model cards reduce in length and standardize by turning, more often, to templates and automatically generated text. Overall, this work takes a step toward an empirically grounded understanding of model fine-tuning and suggests that ecological models and methods can yield novel scientific insights.