Rapid Adaptation with Conditionally Shifted Neurons

We describe a mechanism by which artificial neural networks can learn rapid adaptation - the ability to adapt on the fly, with little data, to new tasks - that we call conditionally shifted neurons. We apply this mechanism in the framework of metalearning, where the aim is to replicate some of the flexibility of human learning in machines. Conditionally shifted neurons modify their activation values with task-specific shifts retrieved from a memory module, which is populated rapidly based on limited task experience. On metalearning benchmarks from the vision and language domains, models augmented with conditionally shifted neurons achieve state-of-the-art results.

[1]  Geoffrey E. Hinton.  Using fast weights to deblur old memories , 1987 .

[2]  Yoshua Bengio,et al.  Learning a synaptic learning rule , 1991, IJCNN-91-Seattle International Joint Conference on Neural Networks.

[3]  Sebastian Thrun,et al.  Explanation-Based Neural Network Learning for Robot Control , 1992, NIPS.

[4]  Jürgen Schmidhuber,et al.  A ‘Self-Referential’ Weight Matrix , 1993 .

[5]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[6]  Sepp Hochreiter,et al.  Learning to Learn Using Gradient Descent , 2001, ICANN.

[7]  Ricardo Vilalta,et al.  A Perspective View and Survey of Meta-Learning , 2002, Artificial Intelligence Review.

[8]  K. Sakai.  Task set and prefrontal cortex. , 2008, Annual review of neuroscience.

[9]  Yoshua Bengio,et al.  Zero-data Learning of New Tasks , 2008, AAAI.

[10]  N. Sigala,et al.  Dynamic Coding for Cognitive Control in Prefrontal Cortex , 2013, Neuron.

[11]  Jürgen Schmidhuber,et al.  Compete to Compute , 2013, NIPS.

[12]  Yoshua Bengio,et al.  An Empirical Investigation of Catastrophic Forgeting in Gradient-Based Neural Networks , 2013, ICLR.

[13]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[14]  Joan Bruna,et al.  Intriguing properties of neural networks , 2013, ICLR.

[15]  Yoshua Bengio,et al.  Towards Biologically Plausible Deep Learning , 2015, ArXiv.

[16]  Markus Siegel,et al.  Cortical information flow during flexible sensorimotor decisions , 2015, Science.

[17]  Earl K. Miller,et al.  Working Memory Capacity: Limits on the Bandwidth of Cognition , 2015, Daedalus.

[18]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[19]  Sanja Fidler,et al.  Predicting Deep Zero-Shot Convolutional Neural Networks Using Textual Descriptions , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[20]  Kenta Oono,et al.  Chainer : a Next-Generation Open Source Framework for Deep Learning , 2015 .

[21]  Joshua B. Tenenbaum,et al.  Human-level concept learning through probabilistic program induction , 2015, Science.

[22]  Phil Blunsom,et al.  Teaching Machines to Read and Comprehend , 2015, NIPS.

[23]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[24]  Gregory R. Koch,et al.  Siamese Neural Networks for One-Shot Image Recognition , 2015 .

[25]  Daan Wierstra,et al.  One-Shot Generalization in Deep Generative Models , 2016, ICML.

[26]  Colin J. Akerman,et al.  Random synaptic feedback weights support error backpropagation for deep learning , 2016, Nature Communications.

[27]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[28]  Jason Weston,et al.  The Goldilocks Principle: Reading Children's Books with Explicit Memory Representations , 2015, ICLR.

[29]  Marcin Andrychowicz,et al.  Learning to learn by gradient descent by gradient descent , 2016, NIPS.

[30]  E. Miller,et al.  Prefrontal Cortex Networks Shift from External to Internal Modes during Learning , 2016, The Journal of Neuroscience.

[31]  Oriol Vinyals,et al.  Matching Networks for One Shot Learning , 2016, NIPS.

[32]  Bartunov Sergey,et al.  Meta-Learning with Memory-Augmented Neural Networks , 2016 .

[33]  Arild Nøkland,et al.  Direct Feedback Alignment Provides Learning in Deep Neural Networks , 2016, NIPS.

[34]  Pieter Abbeel,et al.  Meta-Learning with Temporal Convolutions , 2017, ArXiv.

[35]  Ruslan Salakhutdinov,et al.  Improving One-Shot Learning through Fusing Side Information , 2017, ArXiv.

[36]  Hugo Larochelle,et al.  Optimization as a Model for Few-Shot Learning , 2016, ICLR.

[37]  Hong Yu,et al.  Meta Networks , 2017, ICML.

[38]  Razvan Pascanu,et al.  Overcoming catastrophic forgetting in neural networks , 2016, Proceedings of the National Academy of Sciences.

[39]  Honglak Lee,et al.  Exploring the structure of a real-time, arbitrary neural artistic stylization network , 2017, BMVC.

[40]  Jonathon Shlens,et al.  A Learned Representation For Artistic Style , 2016, ICLR.

[41]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[42]  Philip Bachman,et al.  Learning Algorithms for Active Learning , 2017, ICML.

[43]  Hugo Larochelle,et al.  Modulating early visual processing by language , 2017, NIPS.

[44]  Aaron C. Courville,et al.  FiLM: Visual Reasoning with a General Conditioning Layer , 2017, AAAI.

[45]  G. D. Logan.  Task Switching , 2022 .