BaziQA-Benchmark: Evaluating Symbolic and Temporally Compositional Reasoning in Large Language Models

Jiangxi Chen, Qian Liu

We present BaziQA-Benchmark, a standardized benchmark for evaluating symbolic and temporally compositional reasoning in large language models. The benchmark is derived from 200 pro…