Teaching an agent to perform new tasks using natural language can easily...
Learning from demonstration methods usually leverage close to optimal
de...
When demonstrating a task, human tutors pedagogically modify their behav...
Affordances are the possibilities of actions the environment offers to t...
We introduce SCOD (Sensory Commutativity Object Detection), an active me...
Autonomous discovery and direct instruction are two extreme sources of
l...
We study perception in the scenario of an embodied agent equipped with
f...
In multi-task reinforcement learning there are two main challenges: at
t...
We focus on the problem of teaching a robot to solve tasks presented
seq...
Finding a generally accepted formal definition of a disentangled
represe...
We consider the problem of building a state representation model for con...
Which generative model is the most suitable for Continual Learning? This...
We consider the problem of building a state representation model in a
co...
We propose Flatland, a simple, lightweight environment for fast prototyp...
Skip-gram with negative sampling, a popular variant of Word2vec original...