Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
3 results for “model collapse”
Reviewing Model Collapse and Countermeasures
A new arXiv preprint synthesizes existing research on model collapse — the degradation of AI models trained on synthetic data — to establish foundational understanding, identify mitigation strategies, and outline open challenges.
Aug 25, 2026
The Fairness Collapse Phenomenon: Bias Amplification in Language Models Trained on Synthetic Data
Researchers identify a new phenomenon—'fairness collapse'—where language models trained recursively on synthetic data amplify social biases faster than they degrade in standard performance metrics, posing a stealth risk to AI equity.
Aug 6, 2026
Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning
Researchers propose KITE, a two-stage framework for iterative instruction tuning using synthetic data that aims to prevent model collapse by diagnosing and mitigating competence polarization—where strong skills are reinforced while weak ones degrade—across multiple open-source LLMs.
Jul 21, 2026