Retrieval Threshold Studio

Move a relevance threshold and watch the context set change — a small visual model of the precision-versus-coverage trade-off in retrieval systems.

• Documents and similarity scores are representative fixtures.
• The experiment does not model reranking or query rewriting.
• Score distributions differ significantly between embedding models.

This experiment collects no analytics and stores nothing server-side. Any state you see lives in this browser tab only (memory or localStorage) and clears on reset or reload.

Spotted something broken? Open an issue on the asafarim-platform repo with the experiment slug (retrieval-threshold) in the title.