Explore Reinforced: Equilibrium Approximation with Reinforcement Learning
Ryan Yu 1, Mateusz Nowak 2, Qintong Xie 2, Michelle Yilin Feng 1 and Peter Chin 2 1 - Boston University - Computer Science Department Boston, MA 02215 - United States 2 - Dartmouth College - Thayer School of Engineering Hanover, NH 03755 - United States · 2024
Read Explore Reinforced: Equilibrium Approximation with Reinforcement Learning by Ryan Yu 1, Mateusz Nowak 2, Qintong Xie 2, Michelle Yilin Feng 1 and Peter Chin 2 1 - Boston University - Computer Science Department Boston, MA 02215 - United States 2 - Dartmouth College - Thayer School of Engineering Hanover, NH 03755 - United States (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-02016.txt.
Description
Ryan Yu and co-authors blend in this 2024 paper reinforcement learning with game theory to find equilibria in large, shifting games. Approximation methods for coarse correlated equilibria carry convergence guarantees but choke on big stochastic settings; plain reinforcement learning trains fast yet lands on weaker solutions. Their algorithm, Exp3-IXrl, keeps the learner's action choice apart from the equilibrium math while leaving learning intact.
The split lets equilibrium methods reach environments that were out of bounds before. Tests run in a complex, adversarial cybersecurity gym and in classic multi-armed bandit settings, where the hybrid beat expectations. The authors frame the work as widening where strong solution concepts can actually be computed. Both test beds push the method in unlike ways.
Catalog record
| CREATOR | Ryan Yu 1, Mateusz Nowak 2, Qintong Xie 2, Michelle Yilin Feng 1 and Peter Chin 2 1 - Boston University - Computer Science Department Boston, MA 02215 - United States 2 - Dartmouth College - Thayer School of Engineering Hanover, NH 03755 - United States |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 13 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-02016