# Position: Model Collapse Does Not Mean What You Think

arXiv: https://arxiv.org/abs/2503.03150 · HTML: https://mlsanity.com/paper/2503.03150/

## Key facts

- Cited by 33 papers in the map (data through 14 Oct 2026).
- Placed in Model collapse from synthetic data, one of 13 directions in Synthetic data and self-improvement for LLMs.
- Builds on 44 of its 67 references that are in the map.
- First posted on arXiv 5 Mar 2025.

## Placed in

- Synthetic data and self-improvement for LLMs / [Model collapse from synthetic data](https://mlsanity.com/synth/model-collapse-from-synthetic-data/)

## Builds on (most cited first, 30 of 44)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Analyzing and Improving the Image Quality of StyleGAN](https://mlsanity.com/paper/1912.04958/) (2019)
- [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://mlsanity.com/paper/2403.05530/) (2024)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [Large Scale GAN Training for High Fidelity Natural Image Synthesis](https://mlsanity.com/paper/1809.11096/) (2018)
- [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://mlsanity.com/paper/2101.00027/) (2020)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [Cross-Task Generalization via Natural Language Crowdsourcing Instructions](https://mlsanity.com/paper/2104.08773/) (2021)
- [Dynabench: Rethinking Benchmarking in NLP](https://mlsanity.com/paper/2104.14337/) (2021)
- [DataComp-LM: In search of the next generation of training sets for language models](https://mlsanity.com/paper/2406.11794/) (2024)
- [Fairness Without Demographics in Repeated Loss Minimization](https://mlsanity.com/paper/1806.08010/) (2018)
- [The Curse of Recursion: Training on Generated Data Makes Models Forget](https://mlsanity.com/paper/2305.17493/) (2023)
- [Are GANs Created Equal? A Large-Scale Study](https://mlsanity.com/paper/1711.10337/) (2017)
- [Self-Consuming Generative Models Go MAD](https://mlsanity.com/paper/2307.01850/) (2023)
- [Demographic Dialectal Variation in Social Media: A Case Study of African-American English](https://mlsanity.com/paper/1608.08868/) (2016)
- [Will we run out of data? Limits of LLM scaling based on human-generated data](https://mlsanity.com/paper/2211.04325/) (2022)
- [Does Writing with Language Models Reduce Content Diversity?](https://mlsanity.com/paper/2309.05196/) (2023)
- [Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data](https://mlsanity.com/paper/2404.01413/) (2024)
- [Towards Accountability for Machine Learning Datasets: Practices from Software Engineering and Infrastructure](https://mlsanity.com/paper/2010.13561/) (2020)
- [Predictive Biases in Natural Language Processing Models: A Conceptual Framework and Overview](https://mlsanity.com/paper/1912.11078/) (2019)
- [A Tale of Tails: Model Collapse as a Change of Scaling Laws](https://mlsanity.com/paper/2402.07043/) (2024)
- [Lessons from Archives: Strategies for Collecting Sociocultural Data in Machine Learning](https://mlsanity.com/paper/1912.10389/) (2019)
- [On the Stability of Iterative Retraining of Generative Models on their own Data](https://mlsanity.com/paper/2310.00429/) (2023)
- [Runaway Feedback Loops in Predictive Policing](https://mlsanity.com/paper/1706.09847/) (2017)
- [Model Collapse Demystified: The Case of Regression](https://mlsanity.com/paper/2402.07712/) (2024)
- [Large Language Models Suffer From Their Own Output: An Analysis of the Self-Consuming Training Loop](https://mlsanity.com/paper/2311.16822/) (2023)
- [How Bad is Training on Synthetic Data? A Statistical Analysis of Language Model Collapse](https://mlsanity.com/paper/2404.05090/) (2024)
- [Strong Model Collapse](https://mlsanity.com/paper/2410.04840/) (2024)
- [Data Feedback Loops: Model-driven Amplification of Dataset Biases](https://mlsanity.com/paper/2209.03942/) (2022)

## Built on by (most cited first, 30 of 33)

- [A Closer Look at Model Collapse: From a Generalization-to-Memorization Perspective](https://mlsanity.com/paper/2509.16499/) (2025)
- [When Models Don't Collapse: On the Consistency of Iterative MLE](https://mlsanity.com/paper/2505.19046/) (2025)
- [Rigor in AI: Doing Rigorous AI Work Requires a Broader, Responsible AI-Informed Conception of Rigor](https://mlsanity.com/paper/2506.14652/) (2025)
- [Position: Machine Learning Conferences Should Establish a "Refutations and Critiques" Track](https://mlsanity.com/paper/2506.19882/) (2025)
- [What happens when generative AI models train recursively on each others' outputs?](https://mlsanity.com/paper/2505.21677/) (2025)
- [LLM Web Dynamics: Tracing Model Collapse in a Network of LLMs](https://mlsanity.com/paper/2506.15690/) (2025)
- [Escaping Model Collapse via Synthetic Data Verification: Near-term Improvements and Long-term Convergence](https://mlsanity.com/paper/2510.16657/) (2025)
- [The Impact of AI-Generated Text on the Internet](https://mlsanity.com/paper/2604.26965/) (2026)
- [A Probabilistic Perspective on Model Collapse](https://mlsanity.com/paper/2505.13947/) (2025)
- [Ambiguity Collapse by LLMs: A Taxonomy of Epistemic Risks](https://mlsanity.com/paper/2603.05801/) (2026)
- [Markovian Generation Chains in Large Language Models](https://mlsanity.com/paper/2603.11228/) (2026)
- [Recursive Training Loops in LLMs: How training data properties modulate distribution shift in generated data?](https://mlsanity.com/paper/2504.03814/) (2025)
- [Epistemic diversity across language models mitigates knowledge collapse](https://mlsanity.com/paper/2512.15011/) (2025)
- [The Homogenization Problem in LLMs: Towards Meaningful Diversity in AI Safety](https://mlsanity.com/paper/2601.06116/) (2026)
- [When Does Model Collapse Occur in Structured Interactive Learning?](https://mlsanity.com/paper/2605.20151/) (2026)
- [Losing our Tail, Again: (Un)Natural Selection & Multilingual LLMs](https://mlsanity.com/paper/2507.03933/) (2025)
- [ForTIFAI: Fending Off Recursive Training Induced Failure for AI Model Collapse](https://mlsanity.com/paper/2509.08972/) (2025)
- [First-Extinction Law for Resampling Processes](https://mlsanity.com/paper/2509.20101/) (2025)
- [Synthetic Eggs in Many Baskets: The Impact of Synthetic Data Diversity on LLM Fine-Tuning](https://mlsanity.com/paper/2511.01490/) (2025)
- [Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities](https://mlsanity.com/paper/2605.04127/) (2026)
- [AI-Mediated Communication Can Steer Collective Opinion](https://mlsanity.com/paper/2605.16245/) (2026)
- [Model Collapse as Cultural Evolution](https://mlsanity.com/paper/2605.23054/) (2026)
- [When Sample Selection Bias Precipitates Model Collapse](https://mlsanity.com/paper/2606.13732/) (2026)
- [A Multi-Timescale Recursive Self-Improvement Engine for Open-Ended Persona Growth](https://mlsanity.com/paper/2607.08252/) (2026)
- [The Anti-Ouroboros Effect: Emergent Resilience in Large Language Models from Recursive Selective Feedback](https://mlsanity.com/paper/2509.10509/) (2025)
- [Entropy Minimization without Model Collapse: Mitigating Prediction Bias in Medical Imaging](https://mlsanity.com/paper/2606.02339/) (2026)
- [Generative Models Erode Human Temporal Learning Through Market Selection](https://mlsanity.com/paper/2606.06572/) (2026)
- [Prediction-Only Distillation in Linear and Logistic Regression](https://mlsanity.com/paper/2607.15450/) (2026)
- [An Investigation of the NeurIPS and ICML 2025 Position Tracks](https://mlsanity.com/paper/2608.16894/) (2026)
- [The Oligarch Barely Steers Model Collapse in Multi-Model Ecosystems](https://mlsanity.com/paper/2609.11146/) (2026)

## Abstract (arXiv)

The proliferation of AI-generated content online has fueled concerns over model collapse, a degradation in future generative models' performance when trained on synthetic data generated by earlier models. Industry leaders, premier research journals and popular science publications alike have prophesied catastrophic societal consequences stemming from model collapse. In this position piece, we contend this widespread narrative fundamentally misunderstands the scientific evidence. We highlight that research on model collapse actually encompasses eight distinct and at times conflicting definitions of model collapse, and argue that inconsistent terminology within and between papers has hindered building a comprehensive understanding of model collapse. To assess how significantly different interpretations of model collapse threaten future generative models, we posit what we believe are realistic conditions for studying model collapse and then conduct a rigorous assessment of the literature's methodologies through this lens. While we leave room for reasonable disagreement, our analysis of research studies, weighted by how faithfully each study matches real-world conditions, leads us to conclude that certain predicted claims of model collapse rely on assumptions and conditions that poorly match real-world conditions, and in fact several prominent collapse scenarios are readily avoidable. Altogether, this position paper argues that model collapse has been warped from a nuanced multifaceted consideration into an oversimplified threat, and that the evidence suggests specific harms more likely under society's current trajectory have received disproportionately less attention.