TY  - RPRT
TI  - Evaluating the Paperclip Maximizer: Are RL-Based Language Models More Likely to Pursue Instrumental Goals?
AU  - Yufei He
AU  - Yuexin Li
AU  - Jiaying Wu
AU  - Yuan Sui
AU  - Yulin Chen
AU  - Bryan Hooi
PY  - 2025
UR  - https://arxiv.org/abs/2502.12206
ID  - 2502.12206
ER  -