TY  - RPRT
TI  - Training Language Models to Win Debates with Self-Play Improves Judge Accuracy
AU  - Samuel Arnesen
AU  - David Rein
AU  - Julian Michael
PY  - 2024
UR  - https://arxiv.org/abs/2409.16636
ID  - 2409.16636
ER  -