pyvene: A Library for Understanding and Improving PyTorch Models via Interventions

TL;DR

pyvene is an open-source Python library that enables flexible, customizable interventions on PyTorch models to facilitate research in model editing, interpretability, and robustness, supporting complex intervention schemes with an extensible framework.

Contribution

The paper introduces pyvene, a novel library that simplifies and unifies the process of performing interventions on neural network models in PyTorch, including static and trainable interventions.

Findings

Demonstrates interpretability analyses using causal abstraction and knowledge localization.

Provides a flexible framework for complex interventions on neural models.

Enables sharing and reproducibility of intervened models.

Abstract

Interventions on model-internal states are fundamental operations in many areas of AI, including model editing, steering, robustness, and interpretability. To facilitate such research, we introduce $\textbf{pyvene}$, an open-source Python library that supports customizable interventions on a range of different PyTorch modules. $\textbf{pyvene}$ supports complex intervention schemes with an intuitive configuration format, and its interventions can be static or include trainable parameters. We show how $\textbf{pyvene}$ provides a unified and extensible framework for performing interventions on neural models and sharing the intervened upon models with others. We illustrate the power of the library via interpretability analyses using causal abstraction and knowledge localization. We publish our library through Python Package Index (PyPI) and provide code, documentation, and tutorials at https://github.com/stanfordnlp/pyvene.