APT: Improving Specialist LLM Performance with Weakness Case Acquisition and Iterative Preference Training

Open in new window