Memory Injections: Correcting Multi-Hop Reasoning Failures During Inference in Transformer-Based Language Models

Memory Injections: Correcting Multi-Hop Reasoning Failures During Inference in Transformer-Based Language Models Mansi Sakarvadia author Aswathy Ajith author Arham Khan author Daniel Grzenda author Nathaniel Hudson author André Bauer author Kyle Chard author Ian Foster author 2023-12 text Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP Yonatan Belinkov editor Sophie Hao editor Jaap Jumelet editor Najoung Kim editor Arya McCarthy editor Hosein Mohebbi editor Association for Computational Linguistics Singapore conference publication sakarvadia-etal-2023-memory 10.18653/v1/2023.blackboxnlp-1.26 https://aclanthology.org/2023.blackboxnlp-1.26/ 2023-12 342 356