Discount Model Search for Quality Diversity Optimization in High-Dimensional Measure Spaces

April 25, 2026 | B. Tjanaka, H. Chen, M. Fontaine, and S. Nikolaidis

We propose Discount Model Search (DMS), which guides CMA-MAE exploration with a model that provides a smooth, continuous representation of discount values. DMS works well in high dimensional measure spaces, outperforming CMA-MAE and other existing black-box QD algorithms.

Deep Surrogate Assisted Generation of Environments

June 9, 2022 | V. Bhatt*, B. Tjanaka*, M. C. Fontaine*, S. Nikolaidis

We propose Deep Surrogate Assisted Generation of Environments (DSAGE), a sample-efficient QD environment generation algorithm that maintains a deep surrogate model for predicting agent behaviors in new environments.

Approximating Gradients for Differentiable Quality Diversity in Reinforcement Learning

February 8, 2022 | B. Tjanaka, M. C. Fontaine, J. Togelius, S. Nikolaidis

We propose two variants of the current state-of-the-art DQD algorithm that compute gradients via approximation methods common in reinforcement learning (RL). Evaluation on simulated locomotion tasks indicates our method to achieve comparable performance compared to state-of-the-art.

On the Importance of Environments in Human-Robot Coordination

June 21, 2021 | Matthew C. Fontaine*, Ya-Chuan Hsu* Yulun Zhang*, Bryon Tjanaka, Stefanos Nikolaidis

We demonstrate the effect of the environment on human-robot coordination and introduce a framework for finding environments that elicit different coordination behaviors.