No
Yes
View More
View Less
Working...
Close
OK
Cancel
Confirm
System Message
Delete
Schedule
An unknown error has occurred and your request could not be completed. Please contact support.
Scheduled
Wait Listed
Personal Calendar
Speaking
Conference Event
Meeting
Interest
Schedule TBD
Conflict Found
This session is already scheduled at another time. Would you like to...
Loading...
Please enter a maximum of {0} characters.
Please enter a maximum of {0} words.
must be 50 characters or less.
must be 40 characters or less.
Session Summary
We were unable to load the map image.
This has not yet been assigned to a map.
Search Catalog
Reply
Replies ()
Search
New Post
Microblog
Microblog Thread
Post Reply
Post
Your session timed out.
This web page is not optimized for viewing on a mobile device. Visit this site in a desktop browser to access the full set of features.
2017 GTC San Jose

S7720 - Neural Map: Structured Memory for Deep Reinforcement Learning

Session Speakers
Session Description

We'll address a novel memory system for deep reinforcement learning (DRL) agents. A critical component to enabling intelligent reasoning in partially observable tasks is memory. Despite this importance, DRL agents have so far used simple memory architectures, typically either a temporal convolution over the past k frames or an LSTM layer. We've developed the neural map, which uses a spatially structured 2D memory image along with an adaptable sparse write operator to learn to store information about the environment over long time lags. By making use of the NVIDIA DGX-1 supercomputer, we demonstrate empirically that the neural map surpasses previous DRL memories on a set of challenging 2D and 3D maze environments and show generalization to environments not seen during training.


Additional Session Information
Intermediate
Talk
Algorithms Deep Learning and AI
Higher Education / Research
25 minutes
Session Schedule