LPFQA: A Long-Tail Professional Forum-based Benchmark for LLM Evaluation

Liya Zhu, Peizhuang Cong, Jingzhe Ding +17

Large Language Models (LLMs) perform well on standard reasoning and question-answering benchmarks, yet such evaluations often fail to capture their ability to handle long-tail, exp…