Concept Learning with Energy-Based Models

By Igor Mordatch. 2018

27 citations recorded by the source corpus.

Many hallmarks of human intelligence, such as generalizing from limited experience, abstract reasoning and planning, analogical reasoning, creative problem solving, and capacity for language require the ability to consolidate experience into concepts, which act as basic building blocks of understanding and reasoning. We present a framework that defines a concept by an energy function over events in the environment, as well as an attention mask over entities participating in the event. Given few demonstration events, our method uses inference-time optimization procedure to generate events involving similar concepts or identify entities involved in the concept. We evaluate our framework on learning visual, quantitative, relational, temporal concepts from demonstration events in an unsupervised manner. Our approach is able to successfully generate and identify concepts in a few-shot setting and resulting learned concepts can be reused across environments. Example videos of our results are available at sites.google.com/site/energyconceptmodels

Subfields: Artificial Intelligence

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

Citations in the corpus

EDUCE: Explaining model Decisions through Unsupervised Concepts Extraction (2019, background): “[18] propose a model that learns to deﬁne a concept by a combination of events in an environment.”
On the Binding Problem in Artificial Neural Networks (2020, background): “, 2015), energy-based models (Mordatch, 2019), and capsules (Sabour et al.”
Implicit Behavioral Cloning (2021, background): “221 A general framework for energy-based learning of behaviors is also presented in [34].”
Deep Set Prediction Networks (2019, background): “The closest work to ours is by Mordatch [17].”
Learning a Prior over Intent via Meta-Inverse Reinforcement Learning (2018, methodology): “Our approach builds on work on the broader problem of meta-learning (Schmidhuber, 1987; Bengio et al.; Naik & Mammone, 1992; Thrun & Pratt, 2012) and generative modelling (Rezende et al., 2016; Reed et al., 2018; Mordatch, 2018).”
Set Prediction without Imposing Structure as Conditional Density Estimation (2020, methodology): “Our framework is closely related to the works of Belanger & McCallum (2016) and Mordatch (2018), which also take on an energy-based viewpoint.”
Better Set Representations For Relational Reasoning (2020, methodology): “This idea has been used to predict sets in unsupervised object detection [6], general set prediction [23], and energy-based generative models [14].”
Meta-Learning with Implicit Gradients (2019, methodology): “While we studied different gradient-based optimization methods in the inner loop, iMAML can in principle be used with a variety of inner loop algorithms, including dynamic programming methods such as Q-learning, two-player adversarial games such as GANs, energy-based models [39], and actor-critic RL”