Vector-based navigation using grid-like representations in artificial agents

By Andrea Banino, C. Barry, Benigno Uria, Charles Blundell, Timothy P. Lillicrap, Piotr Wojciech Mirowski, Alexander Pritzel, Martin Chadwick, Thomas Degris, Joseph Modayil and 16 more. 2018

716 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Cognitive Neuroscience, Computer Vision and Pattern Recognition

How researchers see this paper

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

In Aerospace Engineering

In Cognitive Neuroscience

In Computer Vision and Pattern Recognition

Citations in the corpus

If deep learning is the answer, what is the question? (2020, background): “Although machine learning researchers have noted that lattice-like codes may emerge when deep reinforcement learning systems are trained to navigate 113,114 , they have yet to build on these insights for building stronger AI.”
Using Unity to Help Solve Intelligence (2020, background): “For example, DeepMind Lab proved an excellent platform for exploring complex navigational tasks (Banino et al., 2018) and classical laboratory psychological experiments (Leibo et al., 2018), among others, but is limited to ﬁrst-person perspectives and primitive physics.”
The Tolman-Eichenbaum Machine: Unifying Space and Relational Memory through Generalization in the Hippocampal Formation (2019, background): “Following Eichenbaum (14), hippocampal representations, (p), are inferred as a conjunction between sensory input (x) in the lateral (LEC), and abstract location (g) in the medial (MEC) entorhinal cortex.”
What Is a Cognitive Map? Organizing Knowledge for Flexible Behavior. (2018, background): “By using these cells in more complex environments, the agent can solve navigation problems that require vector navigation (such as finding shortcuts that have never previously been taken) (Banino et al., 2018) (Figure 5C) suggesting that this gridlike basis is useful for producing complex behavior.”
Learning to Predict Without Looking Ahead: World Models Without Forward Prediction (2019, neutral): “grid cells) [4, 10].”
Deep neuroethology of a virtual rodent (2019, background): “, 2018) and navigation (Banino et al., 2018; Cueva & Wei, 2018).”
Learning to Navigate in Cities Without a Map (2018, background): “The subject of navigation is attractive to various research disciplines and technology domains alike, being at once a subject of inquiry from the point of view of neuroscientists wishing to crack the code of grid and place cells [3, 13], as well as a fundamental aspect of robotics research.”
Neural network vs. HMM speech recognition systems as models of human cross-linguistic phonetic perception (2018, background): “necessity, several studies have reported a correlation between normative performance of the systems and their ability to predict human behavior or brain activity (e.g. Yamins et al., 2014; Banino et al., 2018).”
Presynaptic stochasticity improves energy efficiency and helps alleviate the stability-plasticity dilemma (2021, background): “As a growing body of literature suggests, artificial neural networks (ANNs) match several aspects of biological neuronal networks in various goal-driven situations (Kriegeskorte, 2015; Yamins and DiCarlo, 2016; Kell et al., 2018; Banino et al., 2018; Cueva and Wei, 2018; Mattar and Daw, 2018).”
Neural belief states for partially observed domains (2018, background): “[4, 5]) often increases performance, but does not capture a clear or interpretable notion of uncertainty.”
The alignment problem from a deep learning perspective (2022, background): “Similarly, Banino et al. [2018] and Wijmans et al. [2023] identiﬁed representations which helped policies plan their routes when navigating, including in unfamiliar settings.”