Completed
- What are these agents really learning?
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Spurious Normativity Enhances Learning of Compliance and Enforcement Behavior in Artificial Agents
Automatically move to the next video in the Classroom when playback concludes
- 1 - Intro
- 2 - Paper Overview
- 3 - Why are some social norms arbitrary?
- 4 - Reinforcement learning environment setup
- 5 - What happens if we introduce a "silly" rule?
- 6 - Experimental Results: how silly rules help society
- 7 - Isolated probing experiments
- 8 - Discussion of the results
- 9 - Start of Interview
- 10 - Where does the research idea come from?
- 11 - What is the purpose behind this research?
- 12 - Short recap of the mechanics of the environment
- 13 - How much does such a closed system tell us about the real world?
- 14 - What do the results tell us about silly rules?
- 15 - What are these agents really learning?
- 16 - How many silly rules are optimal?
- 17 - Why do you have separate weights for each agent?
- 18 - What features could be added next?
- 19 - How sensitive is the system to hyperparameters?
- 20 - How to avoid confirmation bias?
- 21 - How does this play into progress towards AGI?
- 22 - Can we make real-world recommendations based on this?
- 23 - Where do we go from here?
