How Do Large Language Models Understand Relevance? A Mechanistic Interpretability Perspective.

@article{liu-2026-large, author = {Qi Liu and Haozhe Duan and Jiaxin Mao and Ji-Rong Wen}, title = {{How Do Large Language Models Understand Relevance? A Mechanistic Interpretability Perspective}}, journal = {ACM Transactions on Information Systems (TOIS)}, volume = {44}, number = {2}, pages = {34:1--34:27}, year = {2026}, url = {https://doi.org/10.1145/3774942}, doi = {https://doi.org/10.1145/3774942} }