{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/why-does-ctc-result-in-peaky-behavior","title":"Why does CTC result in peaky behavior?","arxiv_id":"2105.14849","date":"2021-05-31","proceeding":null,"authors":["Albert Zeyer","Ralf Schlüter","Hermann Ney"],"abstract":"The peaky behavior of CTC models is well known experimentally. However, an understanding about why peaky behavior occurs is missing, and whether this is a good property. We provide a formal analysis of the peaky behavior and gradient descent convergence properties of the CTC loss and related training criteria. Our analysis provides a deep understanding why peaky behavior occurs and when it is suboptimal. On a simple example which should be trivial to learn for any model, we prove that a feed-forward neural network trained with CTC from uniform initialization converges towards peaky behavior with a 100% error rate. Our analysis further explains why CTC only works well together with the blank label. We further demonstrate that peaky behavior does not occur on other related losses including a label prior model, and that this improves convergence.","url_abs":"https://arxiv.org/abs/2105.14849v2","url_pdf":"https://arxiv.org/pdf/2105.14849v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"why-does-ctc-result-in-peaky-behavior","repo_url":"https://github.com/rwth-i6/returnn-experiments/tree/master/2021-formal-peaky-behavior-ctc","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[{"method_slug":"ctc-loss","method_name":"CTC Loss"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2105.14849","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}