Preprint
Aug 2026
Mitigating LLM sycophancy with RL-based fine-tuning: Bayesian Truth Serum approach
A methodology for mitigating sycophancy that employs the Bayesian Truth Serum (BTS), a peer-prediction mechanism, as the reward in Group Relative Policy Optimization (GRPO) to fine-tune an LLM.
Serhii Mytsyk, Yiming Zhang, Vikram Krishnamurthy
· 0 citations