{ "status": "VALIDATION_SELECTION_LOCKED_BEFORE_FRESH_TEST", "locked_at": "2026-07-15T20:37:11+09:00", "selection_metric": "open_weight_panel_mean_prompt_worst_vs_base", "tie_break": "candidate_id lexical order", "selected_by_method": { "dpo": { "candidate_id": "dpo_b", "validation_primary": 0.40625, "validation_primary_ci95": [ 0.3837890625, 0.427734375 ], "eligible_candidates": [ "dpo_a", "dpo_b" ] }, "ht_mnpo_helpfulness": { "candidate_id": "ht_help_b", "validation_primary": 0.4072265625, "validation_primary_ci95": [ 0.3828125, 0.4296875 ], "eligible_candidates": [ "ht_help_b" ] }, "ht_mnpo_safety": { "candidate_id": "ht_safety_a", "validation_primary": 0.4150390625, "validation_primary_ci95": [ 0.3925537109375, 0.4384765625 ], "eligible_candidates": [ "ht_safety_a", "ht_safety_b" ] }, "inpo_avg": { "candidate_id": "inpo_b", "validation_primary": 0.40625, "validation_primary_ci95": [ 0.3818359375, 0.427734375 ], "eligible_candidates": [ "inpo_b" ] }, "ronpo_full_expect": { "candidate_id": "ronpo_full_a", "validation_primary": 0.4033203125, "validation_primary_ci95": [ 0.37890625, 0.42580566406249987 ], "eligible_candidates": [ "ronpo_full_a", "ronpo_full_b" ] }, "ronpo_k_only": { "candidate_id": "ronpo_top_a", "validation_primary": 0.412109375, "validation_primary_ci95": [ 0.3876953125, 0.43557128906249987 ], "eligible_candidates": [ "ronpo_top_a" ] }, "sppo_avg": { "candidate_id": "sppo_b", "validation_primary": 0.4111328125, "validation_primary_ci95": [ 0.3866943359375, 0.4345703125 ], "eligible_candidates": [ "sppo_b" ] } }, "selected_ronpo_overall": "ronpo_top_a", "failed_methods": [ "ht_mnpo_conciseness", "ipo", "simpo" ], "panel_summary_sha256": "8850f23e16cadf56a99e3d562d224ce5337ab36fa588d8a843382853e6ad17bd", "evaluator_lock_sha256": "ab7b12c84932ed57acfec7edbcbc1ff1df5e9acd203869b1458343adf6e5b2c8", "grid_sha256": "abfd982ef10c3bc71364752241150286ebf06be5e97f087596adf80e55d459e0", "fresh_test_opened": false, "spent_sealed_split_touched": false }