TY  - RPRT
TI  - RL-Obfuscation: Can Language Models Learn to Evade Latent-Space Monitors?
AU  - Rohan Gupta
AU  - Erik Jenner
PY  - 2026
UR  - https://arxiv.org/abs/2506.14261
ID  - 2506.14261
ER  -