KnowledgeDebugger -- an Exploration Tool for Knowledge Localization and Editing in Transformers
For researchers studying knowledge storage and editing in Transformers, this tool lowers the barrier to exploratory analysis, but it is an incremental contribution as it builds on existing libraries.
KnowledgeDebugger is a GUI tool that enables no-code exploration of knowledge localization and editing in Transformers, supporting the initial phase of research by providing interactive access to state-of-the-art editing methods. Case studies demonstrate its effectiveness in replicating recent findings.
Recent research has increasingly focused on understanding how Transformers store and process knowledge, as well as how this knowledge can be edited. Research work in this area is often conducted in two phases: first, phenomena are explored on individual samples. Then, when results appear promising, more statistically robust experiments follow. To support the first phase, we propose KnowledgeDebugger, a GUI-based exploration tool for knowledge localization and editing in Transformers. Our tool - inspired by LM-Debugger - offers no-code access to the methods in EasyEdit, a widely used library of state-of-the-art Knowledge Editing approaches. We demonstrate the tool's effectiveness through case studies of recent findings in this field.