Lyapunov Exponents for Diversity in Differentiable Games

12/24/2021
by   Jonathan Lorraine, et al.
12

Ridge Rider (RR) is an algorithm for finding diverse solutions to optimization problems by following eigenvectors of the Hessian ("ridges"). RR is designed for conservative gradient systems (i.e., settings involving a single loss function), where it branches at saddles - easy-to-find bifurcation points. We generalize this idea to non-conservative, multi-agent gradient systems by proposing a method - denoted Generalized Ridge Rider (GRR) - for finding arbitrary bifurcation points. We give theoretical motivation for our method by leveraging machinery from the field of dynamical systems. We construct novel toy problems where we can visualize new phenomena while giving insight into high-dimensional problems of interest. Finally, we empirically evaluate our method by finding diverse solutions in the iterated prisoners' dilemma and relevant machine learning problems including generative adversarial networks.

READ FULL TEXT

page 7

page 15

page 16

page 17

page 18

page 19

page 20

page 21

research
11/12/2020

Ridge Rider: Finding Diverse Solutions by Following Eigenvectors of the Hessian

Over the last decade, a single algorithm has changed many facets of our ...
research
02/16/2022

How to Fill the Optimum Set? Population Gradient Descent with Harmless Diversity

Although traditional optimization methods focus on finding a single opti...
research
01/22/2022

A Framework to Design Approximation Algorithms for Finding Diverse Solutions in Combinatorial Problems

Finding a single best solution is the most common objective in combinato...
research
07/15/2020

Newton-based Policy Optimization for Games

Many learning problems involve multiple agents optimizing different inte...
research
06/12/2019

Critical Point Finding with Newton-MR by Analogy to Computing Square Roots

Understanding of the behavior of algorithms for resolving the optimizati...
research
04/13/2023

OKRidge: Scalable Optimal k-Sparse Ridge Regression for Learning Dynamical Systems

We consider an important problem in scientific discovery, identifying sp...

Please sign up or login with your details

Forgot password? Click here to reset