Inside the Unfair Judge: A Mechanistic Interpretability Account of LLM-as-Judge Bias

Part of

Advances In Deep Learning Frameworks