Thinking Spaces
Legibility Lab Thinking Spaces are structured environments designed to help readers examine the ideas described in the research papers. They focus on inspection rather than advice.
Thinking Spaces are implemented as GPT environments configured to operate in thinking mode, allowing readers to explore how information is interpreted by large language models and by people.
These environments are designed to operate in thinking mode. In conversational or auto modes, responses may become more directive and less reflective of the structure being examined.
Each Thinking Space is a simple interaction environment where readers can observe how responses form and change as different sources, contexts, or framings enter the interaction.
The goal is not to produce recommendations but to make the structure of interpretation easier to see.
Legibility Lab 