CBEval: A framework for evaluating and interpreting cognitive biases in LLMs

Ammar Shaikh, Raj Abhijit Dandekar, Sreedath Panat +1

Rapid advancements in Large Language models (LLMs) has significantly enhanced their reasoning capabilities. Despite improved performance on benchmarks, LLMs exhibit notable gaps in…