KnowThyself: An Agentic Assistant for LLM Interpretability

TL;DR

KnowThyself is a chat-based tool that simplifies LLM interpretability by integrating model inspection, natural language querying, and visual explanations into an accessible, conversational platform.

Contribution

It introduces a unified, extensible system that consolidates interpretability tools into a user-friendly chat interface with an orchestrator and specialized modules.

Findings

Enhanced accessibility for LLM interpretability

Interactive visualizations and explanations in a chat format

Lowered technical barriers for users

Abstract

We develop KnowThyself, an agentic assistant that advances large language model (LLM) interpretability. Existing tools provide useful insights but remain fragmented and code-intensive. KnowThyself consolidates these capabilities into a chat-based interface, where users can upload models, pose natural language questions, and obtain interactive visualizations with guided explanations. At its core, an orchestrator LLM first reformulates user queries, an agent router further directs them to specialized modules, and the outputs are finally contextualized into coherent explanations. This design lowers technical barriers and provides an extensible platform for LLM inspection. By embedding the whole process into a conversational workflow, KnowThyself offers a robust foundation for accessible LLM interpretability.