{"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/life-long-disentangled-representation","title":"Life-Long Disentangled Representation Learning with Cross-Domain Latent Homologies","arxiv_id":"1808.06508","date":"2018-08-20","proceeding":"NeurIPS 2018 12","authors":["Alessandro Achille","Tom Eccles","Loic Matthey","Christopher P. Burgess","Nick Watters","Alexander Lerchner","Irina Higgins"],"abstract":"Intelligent behaviour in the real-world requires the ability to acquire new\nknowledge from an ongoing sequence of experiences while preserving and reusing\npast knowledge. We propose a novel algorithm for unsupervised representation\nlearning from piece-wise stationary visual data: Variational Autoencoder with\nShared Embeddings (VASE). Based on the Minimum Description Length principle,\nVASE automatically detects shifts in the data distribution and allocates spare\nrepresentational capacity to new knowledge, while simultaneously protecting\npreviously learnt representations from catastrophic forgetting. Our approach\nencourages the learnt representations to be disentangled, which imparts a\nnumber of desirable properties: VASE can deal sensibly with ambiguous inputs,\nit can enhance its own representations through imagination-based exploration,\nand most importantly, it exhibits semantically meaningful sharing of latents\nbetween different datasets. Compared to baselines with entangled\nrepresentations, our approach is able to reason beyond surface-level statistics\nand perform semantically meaningful cross-domain inference.","url_abs":"http://arxiv.org/abs/1808.06508v1","url_pdf":"http://arxiv.org/pdf/1808.06508v1.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":"life-long-disentangled-representation","repo_url":"https://github.com/oliveradk/cult","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1808.06508","atlas_url":"https://app.syntology.ai/?focus=1808.06508","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}