Modular meta-learning

By Ferran Alet, Tomas Lozano-Perez, Leslie Pack Kaelbling. 2018

133 citations recorded by the source corpus.

Many prediction problems, such as those that arise in the context of robotics, have a simplifying underlying structure that, if known, could accelerate learning. In this paper, we present a strategy for learning a set of neural network modules that can be combined in different ways. We train different modular structures on a set of related tasks and generalize to new tasks by composing the learned modules in new ways. By reusing modules to generalize we achieve combinatorial generalization, akin to the "infinite use of finite means" displayed in language. Finally, we show this improves performance in two robotics-related problems.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Meta & Continual Learning

Citations in the corpus

A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms (2019, background): “Alet et al. (2018) proposed a meta learing algorithm to recover a set of specialized modules, but did not establish any connections to causal mechanisms.”
Continual Learning with Recursive Gradient Optimization (2022, background): “…et al., 2018; Serra et al., 2018; Kaushik et al., 2021), add additional components as new tasks arrive (Rusu et al., 2016; Fernando et al., 2017; Alet et al., 2018; Chang et al., 2019; Li et al., 2019), or use larger networks to generate network parameters (Aljundi et al., 2017; Yoon et al.,…”
Wide Neural Networks Forget Less Catastrophically (2021, background): “Expansion-based methods (Rusu et al., 2016; Fernando et al., 2017; Aljundi et al., 2017; Rosenbaum et al., 2018; Chang et al., 2018; Xu and Zhu, 2018; Ferran Alet, 2018; Li et al., 2019; Veniat et al., 2021) add new network modules or experts for new tasks.”
Towards Continual Reinforcement Learning: A Review and Perspectives (2020, background): “More recent work (Devin et al., 2017; Frans et al., 2017; Fernando et al., 2017; Rosenbaum et al., 2017; Meyerson and Miikkulainen, 2017; Ramachandran and Le, 2018; Chang et al., 2018; Liang et al., 2018; Alet et al., 2018; Cases et al., 2019; Yang et al., 2020) has focused on training neural networ”
Efficient Continual Learning with Modular Networks and Task-Driven Priors (2020, background): “For instance, Alet et al. (2018) proposed a modular approach for robotic applications.”
Using Hindsight to Anchor Past Knowledge in Continual Learning (2019, methodology): “Second, modular approaches use different parts of the prediction function for each new task (Fernando et al., 2017; Aljundi et al., 2017; Rosenbaum et al., 2018; Chang et al., 2018; Xu & Zhu, 2018; Ferran Alet, 2018).”
Efficient Lifelong Learning with A-GEM (2018, methodology): “One approach to LLL uses modular compositional models (Fer-nando et al., 2017; Aljundi et al., 2017; Rosenbaum et al., 2018; Chang et al., 2018; Xu & Zhu, 2018; Ferran Alet, 2018), which limit interference among tasks by using different subset of modules for each task.”
RoboNet: Large-Scale Multi-Robot Learning (2019, background): “The multi-task literature [18, 19], lifelong learning literature [20, 21], and meta-learning literature [22, 23] describe ideas that are tightly coupled with this concept.”
Automatically Composing Representation Transformations as a Means for Generalization (2018, background): “…zero-shot generalization is unclear and follow a common paradigm which assumes that the shared representation of monolithic architectures (with the exception of concurrent work [2]) can be shaped by the diversity of tasks in the training distribution as good initializations for future learning.”
Meta-Learning in Neural Networks: A Survey (2020, methodology): “For example, a MAML-based model was shown to work for human pose estimation [195], modular-meta-learning was successfully applied to robotics [130], while a hypernetwork-based model was applied to few-shot clothes ﬁtting for novel fashion items [165].”
Routing Networks and the Challenges of Modular and Compositional Computation (2019, background): “Alet et al. [2018] combine a routing-like approach with other meta-learning approaches, to allow for quick adaptation to new tasks.”
Lifelong Learning of Compositional Structures (2020, background): “Other approaches do not make such an assumption, and instead learn the structure directly from optimization of a cost function [26, 13, 19, 2, 6].”
A large-scale benchmark for few-shot program induction and synthesis (2021, methodology): “To improve this, more compositional methods to meta-learning are increasingly being proposed (Alet et al., 2018; Bengio et al., 2019; Ke et al., 2019; Mendez & Eaton, 2020; Ruis et al., 2020).”
Meta-Learning with Implicit Gradients (2019, background): “Prior work has considered a number of inner loops, ranging from a very general setting where all parameters are adapted using gradient descent [15], to more structured and specialized settings, such as ridge regression [8], Bayesian linear regression [23], and simulated annealing [2].”
Fast and Slow Learning of Recurrent Independent Mechanisms (2021, background): “The most relevant modular meta-learning work is that of Alet et al. (2018), which proposes to learn modular network architecture based on MAML, however their approach relies on pre-trained composable transformations.”
Neurosymbolic Program Synthesis (unknown): “Alternatively, their pretrained weights can be adapted to to each new task [40].”
Graph Element Networks: adaptive, structured computation and memory (2019, methodology): “It is possible to combine gradient-descent with discrete non-local search steps in our optimization process that add or remove nodes from the network, in the style of the BounceGrad algorithm (Alet et al., 2018).”
Multi-Task Learning with Deep Neural Networks: A Survey (2020, background): “In contrast, Modular Meta-Learning learns a computation graph over the modules, meaning that each step of the computation is a discrete composition of a small number of modules, instead of a soft combination of all of them.”
Efficient Lifelong Learning with A-GEM (2018, methodology): “One approach to LLL uses modular compositional models (Fer-nando et al., 2017; Aljundi et al., 2017; Rosenbaum et al., 2018; Chang et al., 2018; Xu & Zhu, 2018; Ferran Alet, 2018), which limit interference among tasks by using different subset of modules for each task.”