{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,11]],"date-time":"2026-03-11T16:39:53Z","timestamp":1773247193378,"version":"3.50.1"},"reference-count":0,"publisher":"MIT Press - Journals","content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Catastrophic forgetting is the notorious vulnerability of neural networks to the changes in the data distribution during learning. This phenomenon has long been considered a major obstacle for using learning agents in realistic continual learning settings. A large body of continual learning research assumes that task boundaries are known during training. However, only a few works consider scenarios in which task boundaries are unknown or not well defined: task-agnostic scenarios. The optimal Bayesian solution for this requires an intractable online Bayes update to the weights posterior.<\/jats:p>\n               <jats:p>We aim to approximate the online Bayes update as accurately as possible. To do so, we derive novel fixed-point equations for the online variational Bayes optimization problem for multivariate gaussian parametric distributions. By iterating the posterior through these fixed-point equations, we obtain an algorithm (FOO-VB) for continual learning that can handle nonstationary data distribution using a fixed architecture and without using external memory (i.e., without access to previous data). We demonstrate that our method (FOO-VB) outperforms existing methods in task-agnostic scenarios. FOO-VB Pytorch implementation is available at https:\/\/github.com\/chenzeno\/FOO-VB.<\/jats:p>","DOI":"10.1162\/neco_a_01430","type":"journal-article","created":{"date-parts":[[2021,9,2]],"date-time":"2021-09-02T22:39:29Z","timestamp":1630622369000},"page":"1-39","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":8,"title":["Task-Agnostic Continual Learning Using Online Variational Bayes with Fixed-Point Updates"],"prefix":"10.1162","author":[{"given":"Chen","family":"Zeno","sequence":"first","affiliation":[{"name":"Department of Electrical Engineering, Technion, Israel Institute of Technology, Haifa 3299993, Israel chenzeno@campus.technion.ac.il"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Itay","family":"Golan","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Technion, Israel Institute of Technology, Haifa 3299993, Israel itaygolan@gmail.com"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Elad","family":"Hoffer","sequence":"additional","affiliation":[{"name":"Habana-Labs, Caesarea 3079821, Israel elad.hoffer@gmail.com"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Daniel","family":"Soudry","sequence":"additional","affiliation":[{"name":"Department of Electrical Engineering, Technion, Israel Institute of Technology, Haifa 3299993, Israel daniel.soudry@gmail.com"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"281","published-online":{"date-parts":[[2021,8,30]]},"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/direct.mit.edu\/neco\/article-pdf\/doi\/10.1162\/neco_a_01430\/1959538\/neco_a_01430.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"http:\/\/direct.mit.edu\/neco\/article-pdf\/doi\/10.1162\/neco_a_01430\/1959538\/neco_a_01430.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,9,2]],"date-time":"2021-09-02T22:39:30Z","timestamp":1630622370000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/doi\/10.1162\/neco_a_01430\/107073\/Task-Agnostic-Continual-Learning-Using-Online"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,8,30]]},"references-count":0,"URL":"https:\/\/doi.org\/10.1162\/neco_a_01430","relation":{},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"value":"0899-7667","type":"print"},{"value":"1530-888X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,8,30]]}}}