{"url":"/method/dhel","slug":"dhel","name":"DHEL","full_name":"Decoupled Hyperspherical Energy Loss","full_name_withheld":false,"description_markdown":"InfoNCE variants demonstrate direct and indirect coupling between the alignment and uniformity terms thus hurting optimisation. The Decoupled Hyperspherical Energy Loss (DHEL) is an NT-Xent variant that completly decouples alignment from uniformity by discarding the corresponding terms from the denominator. In this way optimisation is more efficient and robust to hyper parameter changes.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Bridging Mini-Batch and Asymptotic Analysis in Contrastive Learning: From InfoNCE to Kernel-Based Losses","paper":"/paper/bridging-mini-batch-and-asymptotic-analysis","first_author":"Panagiotis Koromilas","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/bridging-mini-batch-and-asymptotic-analysis"},"source":{"url":"https://arxiv.org/abs/2405.18045v1","title":"Bridging Mini-Batch and Asymptotic Analysis in Contrastive Learning: From InfoNCE to Kernel-Based Losses","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"General","area_id":"general","collection":"Loss Functions","url":"/methods/category/loss-functions","pwc_aliases":[]}],"n_papers_tagged":1,"archive_num_papers":1,"papers_newest_first":[{"paper":"/paper/bridging-mini-batch-and-asymptotic-analysis","title":"Bridging Mini-Batch and Asymptotic Analysis in Contrastive Learning: From InfoNCE to Kernel-Based Losses","date":"2024-05-28","arxiv_id":"2405.18045","n_code_links":1,"syntology":{"ran":9,"of":9,"unverified":0,"pointer_only":9}}],"papers_shown":1,"tasks":[{"task":"/task/contrastive-learning","name":"Contrastive Learning","papers":1},{"task":"/task/representation-learning","name":"Representation Learning","papers":1}],"tasks_shown":2,"n_tasks":2,"usage_by_year":[{"year":"2024","papers":1}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/dhel"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}