HuaSLIM: Human Attention Motivated Shortcut Learning Identification and Mitigation for Large Language models.

Yuqi Ren, Deyi Xiong

Browse the full ACL paper archive.