Martin Tutek, Fateme Hashemi Chaleshtori, Ana Marasovi\'c, Yonatan Belinkov
Measuring Faithfulness of Chains of Thought by Unlearning Reasoning Steps
https://arxiv.org/abs/2502.14829
Measuring Faithfulness of Chains of Thought by Unlearning Reasoning Steps
https://arxiv.org/abs/2502.14829
Comments