MEMIT scales direct memory editing in transformers from single facts to thousands of associations by optimizing MLP weight updates.
How can we know what language models know? Transactions of the Association for Computational Linguistics, 8:423–438
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2022 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Mass-Editing Memory in a Transformer
MEMIT scales direct memory editing in transformers from single facts to thousands of associations by optimizing MLP weight updates.