Editing Models with Task Arithmetic

by   Gabriel Ilharco, et al.

Changing how pre-trained models behave – e.g., improving their performance on a downstream task or mitigating biases learned during pre-training – is a common practice when developing machine learning systems. In this work, we propose a new paradigm for steering the behavior of neural networks, centered around task vectors. A task vector specifies a direction in the weight space of a pre-trained model, such that movement in that direction improves performance on the task. We build task vectors by subtracting the weights of a pre-trained model from the weights of the same model after fine-tuning on a task. We show that these task vectors can be modified and combined together through arithmetic operations such as negation and addition, and the behavior of the resulting model is steered accordingly. Negating a task vector decreases performance on the target task, with little change in model behavior on control tasks. Moreover, adding task vectors together can improve performance on multiple tasks at once. Finally, when tasks are linked by an analogy relationship of the form “A is to B as C is to D", combining task vectors from three of the tasks can improve performance on the fourth, even when no data from the fourth task is used for training. Overall, our experiments with several models, modalities and tasks show that task arithmetic is a simple, efficient and effective way of editing models.


page 27

page 30

page 31


Task Arithmetic in the Tangent Space: Improved Editing of Pre-Trained Models

Task arithmetic has recently emerged as a cost-effective and scalable ap...

CodeEditor: Learning to Edit Source Code with Pre-trained Models

Developers often perform repetitive code editing activities for various ...

SMILE: Self-Distilled MIxup for Efficient Transfer LEarning

To improve the performance of deep learning, mixup has been proposed to ...

TWINS: A Fine-Tuning Framework for Improved Transferability of Adversarial Robustness and Generalization

Recent years have seen the ever-increasing importance of pre-trained mod...

Prompt Algebra for Task Composition

We investigate whether prompts learned independently for different tasks...

Fast Model Editing at Scale

While large pre-trained models have enabled impressive results on a vari...

Model Spider: Learning to Rank Pre-Trained Models Efficiently

Figuring out which Pre-Trained Model (PTM) from a model zoo fits the tar...

Please sign up or login with your details

Forgot password? Click here to reset