{"data":{"slug":"golden-corpus","term":"Golden Corpus","bucket":"data","definition":"Hand-labelled evaluation set with known commercial outcomes. New scoring models must beat the previous generation on it before shipping.","short_definition":"Hand-labelled evaluation set with known commercial outcomes. New scoring models must beat the previous generation on it before shipping.","long_definition":"Golden Corpus is the regression test for scoring. It is a stratified sample of papers with hand-labelled commercial outcomes, signed by a human reviewer. Every new model iteration replays the Golden Corpus and the result is stamped into the Model Iteration record. A model that does not beat the previous generation cannot ship to the daily pipeline.","related_terms":["model-iteration","viability-score"],"related_term_routes":[{"slug":"model-iteration","term":"Model Iteration","route":"/resources/glossary/model-iteration"},{"slug":"viability-score","term":"Viability Score","route":"/resources/glossary/viability-score"}],"canonical_route":"/resources/glossary/golden-corpus","api_route":"/api/v1/resources/glossary/golden-corpus","jsonld_id":"https://sciencetostartup.com/resources/glossary/golden-corpus","variants":[],"tldr":"Hand-labelled evaluation set with known commercial outcomes. New scoring models must beat the previous generation on it before shipping.","key_points":[],"quality_tier":null,"citation_count":null,"source_state":"curated_static","source_module":"apps/web/data/glossary/terms.ts","definition_sections":{"schema_version":1,"intro":"Golden Corpus is the regression test for scoring. It is a stratified sample of papers with hand-labelled commercial outcomes, signed by a human reviewer. Every new model iteration replays the Golden Corpus and the result is stamped into the Model Iteration record. A model that does not beat the previous generation cannot ship to the daily pipeline.","sections":[{"title":"Definition","items":[{"subtitle":"Golden Corpus","text":"Golden Corpus is the regression test for scoring. It is a stratified sample of papers with hand-labelled commercial outcomes, signed by a human reviewer. Every new model iteration replays the Golden Corpus and the result is stamped into the Model Iteration record. A model that does not beat the previous generation cannot ship to the daily pipeline."}]},{"title":"Related vocabulary","items":[{"subtitle":"Model Iteration","text":"One generation of the scoring or prediction model. Each iteration carries a version, training data hash, and Golden Corpus benchmark."},{"subtitle":"Viability Score","text":"A 0-10 rating of an AI paper's commercial viability. Higher scores cluster with code, demo, dataset, and beats-SOTA signals."}]}],"cited_arxiv_ids":[]}},"meta":{"canonical_route":"/resources/glossary/golden-corpus","api_route":"/api/v1/resources/glossary/golden-corpus","source":{"label":"curated glossary catalog","source_state":"curated_static","source_module":"apps/web/data/glossary/terms.ts","method_version":"public_glossary_curated_terms_v2","freshness":{"status":"versioned","observed_at":null,"fresh_until":null,"reason":"Git-versioned curated catalog; daily ingestion freshness windows do not apply.","reason_code":"git_versioned_curated_catalog"},"source_count":111,"bucket_count":7,"buckets":["scoring","surfaces","agents","distribution","data","foresight","buildability"]}}}