MEND: Fast Model Editing at Scale
MEND: Fast Model Editing at Scale
A computationally efficient approach for learning to edit the behavior of very large pre-trained language models (10 billion+ parameters)
Embedded Files
Abstract
Abstract
While large pre-trained models have enabled impressive results on a variety of downstream tasks, the largest existing models still make errors, and even accurate predictions may become outdated over time. Because detecting all such failures at training time is impossible, enabling both developers and end users of such models to correct inaccurate outputs while leaving the model otherwise intact is desirable. However, the distributed, black-box nature of the representations learned by large neural networks makes producing such targeted edits difficult. If presented with only a single problematic input and new desired output, fine-tuning approaches tend to overfit; other editing algorithms are either computationally infeasible or simply ineffective when applied to very large models. To enable easy post-hoc editing at scale, we propose Model Editor Networks with Gradient Decomposition (MEND), a collection of small auxiliary editing networks that use a single desired input-output pair to make fast, local edits to a pre-trained model. MEND learns to transform the gradient obtained by standard fine-tuning, using a low-rank decomposition of the gradient to make the parameterization of this transformation tractable. MEND can be trained on a single GPU in less than a day even for 10 billion+ parameter models; once trained MEND enables rapid application of new edits to the pre-trained model. Our experiments with T5, GPT, BERT, and BART models show that MEND is the only approach to model editing that produces effective edits for models with tens of millions to over 10 billion parameters.
MEND learns to transform the raw fine-tuning gradient for a weight matrix into a new update that edits the model's behavior while minimizing model degradation on other inputs.
If this paper or code was helpful for your research, please use the following citation:
@article{mitchell2021fast,
title={Fast Model Editing at Scale},
author={Mitchell, Eric and Lin, Charles and Bosselut, Antoine and Finn, Chelsea and Manning, Christopher D.},
year={2021},
journal={CoRR},
url={https://arxiv.org/pdf/2110.11309.pdf}
}
Page updated
Google Sites
Report abuse