Geometric Exploration, Arithmetic Exploitation — LessWrong
This post is going to mostly be propaganda for Thompson sampling. However, the presentation is quite different from the standard presentation. I will…
x Geometric Exploration, Arithmetic Exploitation — LessWrong Geometric Rationality Rationality World Modeling World Optimization Frontpage 142 Geometric Exploration, Arithmetic Exploitation by Scott Garrabrant 24th Nov 2022 8 min read 5 142 This post is going to mostly be propaganda for Thompson sampling. However, the presentation is quite different from the standard presentation. I will be working within a toy model, but I think some of the lessons will generalize. I end with some discussion of fairness and exploitation. A Exploration/Exploitation Toy Model Imagine you are interacting with th
Explore this link on the map →related reading
- The Multi-Armed Bandit Problem and Its Solutions | Lil'Loglilianweng.github.io
- The Bayes Banditfrancesco215.github.io
- On The Independence Axiom — LessWronglesswrong.com
- An Intuitive Explanation of Solomonoff Induction — LessWronglesswrong.com
- Competing with sampling — Alignment Research Centeralignment.org
- The Geometric Expectation — LessWronglesswrong.com
- ARC progress update: Competing with sampling — LessWronglesswrong.com
- What is AIXI?jan.leike.name
- Approximately Bayesian Reasoning: Knightian Uncertainty, Goodhart, and the Look-Elsewhere Effect — LessWronglesswrong.com
- [1902.09469] Embedded Agencyarxiv.org
- The Stamp Collectormindingourway.com
- Geometric UDT — LessWronglesswrong.com