LLM Attributor: Interactive Visual Attribution for LLM Generation

TL;DR

LLM Attributor is an interactive visualization tool that helps users understand and inspect the training data influences on large language model outputs, improving transparency and trust.

Contribution

The paper introduces LLM Attributor, a novel Python library that provides interactive visualizations for data attribution in LLMs, supporting model inspection and comparison.

Findings

Enables visualization of training data influence on LLM outputs

Supports integration with computational notebooks for ease of use

Demonstrated on LLaMA2 models with diverse datasets

Abstract

While large language models (LLMs) have shown remarkable capability to generate convincing text across diverse domains, concerns around its potential risks have highlighted the importance of understanding the rationale behind text generation. We present LLM Attributor, a Python library that provides interactive visualizations for training data attribution of an LLM's text generation. Our library offers a new way to quickly attribute an LLM's text generation to training data points to inspect model behaviors, enhance its trustworthiness, and compare model-generated text with user-provided text. We describe the visual and interactive design of our tool and highlight usage scenarios for LLaMA2 models fine-tuned with two different datasets: online articles about recent disasters and finance-related question-answer pairs. Thanks to LLM Attributor's broad support for computational notebooks, users can easily integrate it into their workflow to interactively visualize attributions of their models. For easier access and extensibility, we open-source LLM Attributor at https://github.com/poloclub/ LLM-Attribution. The video demo is available at https://youtu.be/mIG2MDQKQxM.