{"id":"arxiv:2312.10549","title":"Catastrophic Forgetting in Deep Learning: A Comprehensive Taxonomy","date":"2023-12-16","section":"Data, training, optimization","quality":-0.651376,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/","source_url":"https://arxiv.org/abs/2312.10549","text_available":true,"authors":["Everton L. Aleixo","Juan G. Colonna","Marco Cristo","Everlandio Fernandes"],"outgoing":[],"incoming":[],"text_parts":[{"start":0,"end":15291,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html"},{"start":15291,"end":31008,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-2.html"},{"start":31008,"end":46529,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-3.html"},{"start":46529,"end":62020,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-4.html"},{"start":62020,"end":77087,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-5.html"},{"start":77087,"end":93041,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-6.html"},{"start":93041,"end":108678,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-7.html"},{"start":108678,"end":123878,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-8.html"},{"start":123878,"end":132713,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-9.html"}],"contents":[{"title":"Catastrophic Forgetting in Deep Learning: A Comprehensive Taxonomy","offset":423,"level":1,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"Abstract","offset":493,"level":6,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"1 Introduction","offset":1570,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"2 Related work","offset":8328,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"3 Taxonomic Organization","offset":13103,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"3.1 Rehearsal","offset":14456,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"3.1.1 Pseudo-rehearsal","offset":15262,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-1.html#text"},{"title":"3.1.2 Mini-rehearsal","offset":27181,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-2.html#text"},{"title":"3.2 Distance-based","offset":46505,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-3.html#text"},{"title":"3.3 Sub-networks","offset":52450,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-4.html#text"},{"title":"3.3.1 Soft Sub-networks","offset":57435,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-4.html#text"},{"title":"3.3.2 Hard Sub-networks","offset":71520,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-5.html#text"},{"title":"3.4 Dynamic Networks","offset":81990,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-6.html#text"},{"title":"3.5 Hybrid Approaches","offset":96462,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-7.html#text"},{"title":"3.5.1 Enhancing Distance-based using Rehearsal","offset":97099,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-7.html#text"},{"title":"3.5.2 Enhancing Distance-based using Dynamic Networks","offset":103151,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-7.html#text"},{"title":"3.5.3 Enhancing Dynamic Networks using Rehearsal","offset":104678,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-7.html#text"},{"title":"3.5.4 Enhancing Sub-networks using Rehearsal","offset":111028,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-8.html#text"},{"title":"3.5.5 Growth Sub-networks using Dynamic Networks","offset":113979,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-8.html#text"},{"title":"3.5.6 Combining Sub-networks, Dynamic Networks, and Rehearsal","offset":118453,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-8.html#text"},{"title":"4 Final Considerations","offset":120145,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/part-8.html#text"}],"text_code_points":132713,"source_markdown":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/890b8ca392bd314b377271caffd70091a482d479f8cc7747c31633479b2e87d8/source.md"}