Open access
Jul 2026
CRiT-QA: Evaluating Multi-hop Reasoning with Counterfactual Chains and Distractor Traps
The introduction of CRiT-QA (Counterfactual Reasoning with Traps), a dataset explicitly designed to address both limitations of large language models' multi-hop reasoning, and provides a foundation for developing more reliable, evidence-grounded LLMs.
Jungmin Yun, Junehyoung Kwon, Youngbin Kim
· Proceedings of the Language... · 0 citations