InTraVisTo: Inside Transformer Visualisation Tool
This tool addresses the problem of interpretability for researchers and practitioners working with complex LLMs, though it is incremental as it builds on existing visualization techniques.
The authors tackled the challenge of understanding unpredictable behaviors in large language models (LLMs) by developing InTraVisTo, a tool that visualizes internal states and information flow in Transformer-based models to trace token generation processes.
The reasoning capabilities of Large Language Models (LLMs) have increased greatly over the last few years, as have their size and complexity. Nonetheless, the use of LLMs in production remains challenging due to their unpredictable nature and discrepancies that can exist between their desired behavior and their actual model output. In this paper, we introduce a new tool, InTraVisTo (Inside Transformer Visualisation Tool), designed to enable researchers to investigate and trace the computational process that generates each token in a Transformer-based LLM. InTraVisTo provides a visualization of both the internal state of the Transformer model (by decoding token embeddings at each layer of the model) and the information flow between the various components across the different layers of the model (using a Sankey diagram). With InTraVisTo, we aim to help researchers and practitioners better understand the computations being performed within the Transformer model and thus to shed some light on internal patterns and reasoning processes employed by LLMs.