The information geometry of large language models is shared, learned, and controllable
Fisher-Rao output geometry is shared across LLM architectures and enables minimum-disturbance interventions that transfer across prompts for steering and editing.
1 stories, newest first · one card per story
Fisher-Rao output geometry is shared across LLM architectures and enables minimum-disturbance interventions that transfer across prompts for steering and editing.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views