TY  - RPRT
TI  - Information based explanation methods for deep learning agents -- with applications on large open-source chess models
AU  - Patrik Hammersborg
AU  - Inga Strümke
PY  - 2023
UR  - https://arxiv.org/abs/2309.09702
ID  - 2309.09702
ER  -