Skip to content
#

options-framework

Here are 14 public repositories matching this topic...

Reinforcement learning with options built by construction, not discovery: NetHack's raw keystrokes become a catalogue of closed-loop options enumerated from the action set alone. Same PPO agent, same 10M-step budget, roughly 8x the return of acting over primitives.

  • Updated Sep 25, 2026
  • Python

Add this topic to your repo

To associate your repository with the options-framework topic, visit your repo's landing page and select "manage topics."

Learn more