Papers of the day   All papers

Learning to Continually Learn

Comments

Jeff Clune: Learning to Continuously Learn. ANML meta-learns to reduce catastrophic forgetting, and can learn at least 600 tasks (Omniglot classes) sequentially and performs well on most afterwards. arXiv https://arxiv.org/abs/2002.09571 NeurIPS talk https://slideslive.com/38924020/how-alternatives-to-conventional-neural-networks-loosely-inspired-by-biology Led by Shawn Beaulieu 1/2 https://t.co/8gNm8jCOLE

6 replies, 336 likes


Jeff Clune: Here is a new talk entirely on ANML (a Neuromodulated Meta-Learning Algorithm) @reworkAI ANML meta-learns to reduce catastrophic forgetting, and can learn at least 600 tasks sequentially! http://videos.re-work.co/videos/1861-anml-learning-to-continually-learn paper: Learning to Continually Learn (https://arxiv.org/abs/2002.09571) #AI https://t.co/kea4jA412i

8 replies, 202 likes


roadrunner01: Learning to Continually Learn pdf: https://arxiv.org/pdf/2002.09571.pdf abs: https://arxiv.org/abs/2002.09571 https://t.co/BD6dod4szJ

0 replies, 45 likes


Jeff Clune: AI-GAs: https://arxiv.org/abs/1905.10985 GTNs: https://eng.uber.com/generative-teaching-networks ANML: https://arxiv.org/abs/2002.09571 LRL: https://arxiv.org/abs/1611.05763 RL^2: https://arxiv.org/abs/1611.02779 Reversible Learning (via hypergradients): https://arxiv.org/pdf/1502.03492.pdf

1 replies, 40 likes


Vermont Complex Systems Center @ UVM: "Learning to Continually Learn” New study led by PhD student @_SalvorHardin and faculty member Nick Cheney w/@jeffclune https://arxiv.org/abs/2002.09571 https://t.co/DPky4xnKJC

0 replies, 23 likes


Jeff Clune: PS. The code is available https://github.com/uvm-neurobotics-lab/ANML as is the publication: https://arxiv.org/pdf/2002.09571.pdf

0 replies, 7 likes


arxiv: Learning to Continually Learn. http://arxiv.org/abs/2002.09571 https://t.co/D5BVhTbAPm

0 replies, 3 likes


David Ticoll: Lifelong #machinelearning? "We advocate a solution to catastrophic forgetting, allowing AI to learn to continually learn. Inspired by processes in the brain it enables context-dependent selective activation within a deep neural network." ⁦@jeffclune⁩ https://arxiv.org/abs/2002.09571

0 replies, 2 likes


AI Papers: Learning to Continually Learn. http://arxiv.org/abs/2002.09571

0 replies, 1 likes


Brundage Bot: Learning to Continually Learn. Shawn Beaulieu, Lapo Frati, Thomas Miconi, Joel Lehman, Kenneth O. Stanley, Jeff Clune, and Nick Cheney http://arxiv.org/abs/2002.09571

1 replies, 1 likes


Content

Found on Feb 25 2020 at https://arxiv.org/pdf/2002.09571.pdf

PDF content of a computer science paper: Learning to Continually Learn