Jul 2026
CASE: Causal Alignment and Structural Enforcement for Improving Chain-of-Thought Faithfulness
Experiments show that CASE achieves a 37\% average per-setting relative improvement in overall CoT faithfulness over the strongest baselines, exhibits stronger cross-dataset faithfulness transfer, and maintains competitive average accuracy.
Ziming Wang, Yinghua Yao, Changwu Huang et al.
· arXiv.org · 0 citations