1 paper · 1 filter
Tianyi Huang, Caden Yang, Emily Yin +2
Retrieval-augmented language models can retrieve relevant evidence yet still commit to answers before explicitly checking whether the retrieved context supports the conclusion. We…