DuPO: Enabling Reliable LLM Self-Verification via Dual Preference Optimization
FreeReliable LLM self-verification through dual preference optimization
FreeFree tier
About DuPO: Enabling Reliable LLM Self-Verification via Dual Preference Optimization
DuPO is a method for enabling reliable self-verification in large language models through dual preference optimization, as described in the research paper available on arXiv.