MT-Bench
Emerging6papers using it
2024first seen
MT-Bench is a preference dataset used to evaluate the performance of language model alignment methods by comparing chosen and rejected responses.
Papers using MT-Bench (6)
- SimPER: A Minimalist Approach to Preference Alignment without
HyperparametersPhi-3 Technical Report: A Highly Capable Language Model Locally on Your
PhoneIs In-Context Learning Sufficient for Instruction Following in LLMs?Xwin-LM: Strong and Scalable Alignment Practice for LLMsStable Code Technical ReportI-SHEEP: Self-Alignment of LLM from Scratch through an Iterative
Self-Enhancement Paradigm