karenina.storage.views.template_results¶
template_results
¶
template_results_view
Primary view for verification outcomes. One row per result with pass/fail status, result_id, run_id, run_name, question_id, and benchmark_name. Use for pass rate analysis, filtering failed questions, and joining to other result-level views.
Columns
result_id (TEXT): Unique identifier for the verification result run_id (TEXT): Unique identifier for the verification run (UUID) verification_date (TIMESTAMP): Timestamp when the verification was performed run_name (TEXT): Name of the verification run benchmark_name (TEXT): Name of the benchmark question_id (TEXT): Unique identifier for the question (MD5 hash) question_text (TEXT): The question content verify_result (BOOLEAN): Verification outcome (true=pass, false=fail, NULL=not evaluated) replicate (INTEGER): Replicate number (NULL for single runs, 1/2/3/... for replicated)
Keys
Primary: result_id Joins: result_id → results_metadata_view, raw_llm_answers_view, rubric_traits_view, result_mcp_servers_view run_id → combination_info_view.run_id run_name → combination_info_view.run_name question_id → question_attributes_view.question_id
Example
SELECT run_name, COUNT() as total, SUM(verify_result) as passed, ROUND(100.0 * SUM(verify_result) / COUNT(), 2) as pass_rate FROM template_results_view GROUP BY run_name;