TY  - RPRT
TI  - ApproxED: Approximate exploitability descent via learned best responses
AU  - Carlos Martin
AU  - Tuomas Sandholm
PY  - 2024
UR  - https://arxiv.org/abs/2301.08830
ID  - 2301.08830
ER  -