search

OREL-group / Admin

Administration & Lab Management
1 stars 0 forks source link

RL agent Switching #48

Open jesparent opened 2 years ago

jesparent commented 2 years ago

In this project we aim to show how Switching behavior influences RL agents representation , and whether that helps them learn better , i.e , become more adaptable. This is primarily done by comparing an Agent left for X epochs in a single environment to an Agent left for (x+y+x*=X) epochs in 2 environments x and y .

jesparent commented 2 years ago

@Agrover112 (if this is your project - I wasn't sure of the title used in @balicea 's spreadsheet), please edit the description field above or make a comment here describing your project idea in more detail. Thanks!

Agrover112 commented 2 years ago

In this project we aim to show how Switching behavior influences RL agents representation , and whether that helps them learn better , i.e , become more adaptable. This is primarily done by comparing an Agent left for X epochs in a single environment to an Agent left for (x+y+x*=X) epochs in 2 environments x and y .