factually
Index/Topics/Human-Preference Tuning

Human-Preference Tuning

The process of fine-tuning language models based on human preferences, which can lead to sycophancy.

Fact-Checks

1 result
Jan 22, 2026
Most Viewed

Which AI/LLM has the lowest sycophancy?

— a cross-model evaluation introduced by researchers — measured sycophantic behavior across multiple large language models and found to have the lowest sycophancy score (56.71%) while scored highest (...