{"slug": "mitigating-early-training-collapse-in-ctr-models", "title": "Mitigating Early Training Collapse in CTR Models", "summary": "Researchers studying click-through rate prediction models found that controlling feature sparsity, rather than reducing learning rates, effectively mitigates early training collapse—a sharp validation performance drop after the first epoch. Removing highly sparse features and aggregating infrequent values stabilized training, improved offline metrics, and boosted online system performance on large-scale industrial datasets.", "body_md": "# Computer Science > Machine Learning\n\n[Submitted on 20 Jun 2026]\n\n# Title:Mitigating Early Training Collapse in CTR Models\n\n[View PDF](/pdf/2607.09696)\n\n[HTML (experimental)](https://arxiv.org/html/2607.09696v1)\n\nAbstract:Deep neural models for click-through rate prediction often exhibit a sharp decline in validation performance immediately after the first training epoch despite continued improvement in training loss. This instability restricts effective learning and limits model performance. In this study, we analyze this behavior using large-scale industrial datasets and evaluate practical mitigation strategies. While reducing the learning rate provides only incremental gains, controlling feature sparsity yields substantial improvements. Removing highly sparse features and aggregating infrequent feature values stabilizes training, extends useful learning beyond a single epoch, and improves both offline evaluation metrics and online system performance.\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer\n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers\n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps\n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations\n\n*(*[What are Smart Citations?](https://www.scite.ai/))# Code, Data and Media Associated with this Article\n\nalphaXiv\n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers\n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub\n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub\n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face\n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast\n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower\n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender\n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\nIArxiv Recommender\n\n*(*[What is IArxiv?](https://iarxiv.org/about))# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [ Learn more about arXivLabs](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/mitigating-early-training-collapse-in-ctr-models", "canonical_source": "https://arxiv.org/abs/2607.09696", "published_at": "2026-07-14 04:00:00+00:00", "updated_at": "2026-07-14 04:22:37.518723+00:00", "lang": "en", "topics": ["machine-learning", "artificial-intelligence"], "entities": [], "alternates": {"html": "https://wpnews.pro/news/mitigating-early-training-collapse-in-ctr-models", "markdown": "https://wpnews.pro/news/mitigating-early-training-collapse-in-ctr-models.md", "text": "https://wpnews.pro/news/mitigating-early-training-collapse-in-ctr-models.txt", "jsonld": "https://wpnews.pro/news/mitigating-early-training-collapse-in-ctr-models.jsonld"}}