{"schema":"stdetail-evaluation-bundle-v1","title":"Synthetic HER2 evaluation","workflow":"her2","synthetic":true,"metrics":[{"id":"model_accuracy","label":"Model ordering accuracy","description":"Saved answer accuracy on source-defined, judge-half RNA questions.","unit":"accuracy","higher_is_better":true,"object":"prediction"},{"id":"accuracy_relative_to_rna_repeatability","label":"Accuracy relative to RNA repeatability","description":"(Model accuracy minus held-out gene-majority accuracy) divided by (RNA accuracy minus held-out gene-majority accuracy).","unit":"score","higher_is_better":true,"object":"prediction"},{"id":"lopo_gene_majority_baseline","label":"Held-out gene-majority baseline","description":"Gene-majority answer accuracy fitted using other patients.","unit":"accuracy","higher_is_better":true,"object":"baseline"},{"id":"rna_split_half_accuracy","label":"RNA split-half accuracy","description":"RNA reference answer accuracy on the same source-defined questions.","unit":"accuracy","higher_is_better":true,"object":"rna_reference"}],"rows":[{"model":"aligned_example","task":"Across regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"model_accuracy","value":0.9816319444444443,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9816319444444443,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.46988219246031737,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":1.0121161289288605,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Across regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":1.0121161289288605,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9816319444444443,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.46988219246031737,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":1.0121161289288605,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"Gene-majority baseline","task":"Across regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"lopo_gene_majority_baseline","value":0.511749751984127,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9816319444444443,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.46988219246031737,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":1.0121161289288605,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"baseline"},{"model":"RNA reference","task":"Across regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"rna_split_half_accuracy","value":0.9760069444444444,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9816319444444443,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.46988219246031737,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":1.0121161289288605,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"rna_reference"},{"model":"permuted_example","task":"Across regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"model_accuracy","value":0.5139397321428572,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5139397321428572,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.002189980158730176,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":0.004717170125301539,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"permuted_example","task":"Across regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":0.004717170125301539,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5139397321428572,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.002189980158730176,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":0.004717170125301539,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Across regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"model_accuracy","value":0.9828125,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9828125,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":0.45572916666666663,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":1.0,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Across regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":1.0,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9828125,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":0.45572916666666663,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":1.0,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"Gene-majority baseline","task":"Across regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"lopo_gene_majority_baseline","value":0.5270833333333333,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9828125,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":0.45572916666666663,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":1.0,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"baseline"},{"model":"RNA reference","task":"Across regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"rna_split_half_accuracy","value":0.9828125,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9828125,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":0.45572916666666663,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":1.0,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"rna_reference"},{"model":"permuted_example","task":"Across regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"model_accuracy","value":0.4924627976190476,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4924627976190476,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":-0.03462053571428575,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":-0.07596734693877559,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"permuted_example","task":"Across regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":-0.07596734693877559,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4924627976190476,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":-0.03462053571428575,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":-0.07596734693877559,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Context and depth matched","unit_id":"synthetic_P1","unit_type":"patient","metric":"model_accuracy","value":0.9795833333333333,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9795833333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.47764632936507934,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":1.002576890240093,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"aligned_example","task":"Context and depth matched","unit_id":"synthetic_P1","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":1.002576890240093,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9795833333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.47764632936507934,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":1.002576890240093,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"Gene-majority baseline","task":"Context and depth matched","unit_id":"synthetic_P1","unit_type":"patient","metric":"lopo_gene_majority_baseline","value":0.5019370039682539,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9795833333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.47764632936507934,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":1.002576890240093,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"baseline"},{"model":"RNA reference","task":"Context and depth matched","unit_id":"synthetic_P1","unit_type":"patient","metric":"rna_split_half_accuracy","value":0.9783556547619048,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9795833333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.47764632936507934,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":1.002576890240093,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"rna_reference"},{"model":"permuted_example","task":"Context and depth matched","unit_id":"synthetic_P1","unit_type":"patient","metric":"model_accuracy","value":0.5088070436507937,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5088070436507937,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.006870039682539741,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":0.014420173666784787,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"permuted_example","task":"Context and depth matched","unit_id":"synthetic_P1","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":0.014420173666784787,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5088070436507937,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.006870039682539741,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":0.014420173666784787,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"aligned_example","task":"Context and depth matched","unit_id":"synthetic_P2","unit_type":"patient","metric":"model_accuracy","value":0.98046875,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.98046875,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":0.48735615079365074,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":0.9996794953323327,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"aligned_example","task":"Context and depth matched","unit_id":"synthetic_P2","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":0.9996794953323327,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.98046875,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":0.48735615079365074,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":0.9996794953323327,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"Gene-majority baseline","task":"Context and depth matched","unit_id":"synthetic_P2","unit_type":"patient","metric":"lopo_gene_majority_baseline","value":0.49311259920634926,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.98046875,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":0.48735615079365074,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":0.9996794953323327,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"baseline"},{"model":"RNA reference","task":"Context and depth matched","unit_id":"synthetic_P2","unit_type":"patient","metric":"rna_split_half_accuracy","value":0.9806250000000001,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.98046875,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":0.48735615079365074,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":0.9996794953323327,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"rna_reference"},{"model":"permuted_example","task":"Context and depth matched","unit_id":"synthetic_P2","unit_type":"patient","metric":"model_accuracy","value":0.4838020833333333,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4838020833333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":-0.009310515873015957,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":-0.019098008292422523,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"permuted_example","task":"Context and depth matched","unit_id":"synthetic_P2","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":-0.019098008292422523,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4838020833333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":-0.009310515873015957,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":-0.019098008292422523,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},"model_role":"prediction"},{"model":"aligned_example","task":"Within regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"model_accuracy","value":0.9909722222222221,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9909722222222221,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":0.46185763888888887,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":1.0013550645537697,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Within regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":1.0013550645537697,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9909722222222221,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":0.46185763888888887,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":1.0013550645537697,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"Gene-majority baseline","task":"Within regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"lopo_gene_majority_baseline","value":0.5291145833333333,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9909722222222221,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":0.46185763888888887,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":1.0013550645537697,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"baseline"},{"model":"RNA reference","task":"Within regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"rna_split_half_accuracy","value":0.9903472222222222,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9909722222222221,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":0.46185763888888887,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":1.0013550645537697,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"rna_reference"},{"model":"permuted_example","task":"Within regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"model_accuracy","value":0.5208506944444444,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5208506944444444,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":-0.008263888888888848,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":-0.017916964655399468,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"permuted_example","task":"Within regions","unit_id":"synthetic_P1","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":-0.017916964655399468,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5208506944444444,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":-0.008263888888888848,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":-0.017916964655399468,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Within regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"model_accuracy","value":0.9817708333333334,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9817708333333334,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":0.44848958333333333,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":1.0034960960261041,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"aligned_example","task":"Within regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":1.0034960960261041,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9817708333333334,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":0.44848958333333333,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":1.0034960960261041,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"Gene-majority baseline","task":"Within regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"lopo_gene_majority_baseline","value":0.53328125,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9817708333333334,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":0.44848958333333333,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":1.0034960960261041,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"baseline"},{"model":"RNA reference","task":"Within regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"rna_split_half_accuracy","value":0.9802083333333333,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9817708333333334,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":0.44848958333333333,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":1.0034960960261041,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"rna_reference"},{"model":"permuted_example","task":"Within regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"model_accuracy","value":0.4523958333333333,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4523958333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":-0.08088541666666677,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":-0.18098123761799348,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"},{"model":"permuted_example","task":"Within regions","unit_id":"synthetic_P2","unit_type":"patient","metric":"accuracy_relative_to_rna_repeatability","value":-0.18098123761799348,"status":"ESTIMABLE","support":"source-defined, judge-half questions on common model support","details":{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4523958333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":-0.08088541666666677,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":-0.18098123761799348,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},"model_role":"prediction"}],"sources":[{"name":"run1/patient_scores.tsv","columns":["arena","patient_id","model_id","model_accuracy","seeds_evaluable","lopo_gene_majority_baseline","rna_split_half_accuracy","model_increment","rna_repeatable_increment","accuracy_relative_to_rna_repeatability","status","seeds_requested","task"],"rows":[{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9816319444444443,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.46988219246031737,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":1.0121161289288605,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5139397321428572,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.511749751984127,"rna_split_half_accuracy":0.9760069444444444,"model_increment":0.002189980158730176,"rna_repeatable_increment":0.4642571924603174,"accuracy_relative_to_rna_repeatability":0.004717170125301539,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9828125,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":0.45572916666666663,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":1.0,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},{"arena":"BROAD_STRUCTURE_ALLOWED","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4924627976190476,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5270833333333333,"rna_split_half_accuracy":0.9828125,"model_increment":-0.03462053571428575,"rna_repeatable_increment":0.45572916666666663,"accuracy_relative_to_rna_repeatability":-0.07596734693877559,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Across regions"},{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9795833333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.47764632936507934,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":1.002576890240093,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5088070436507937,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5019370039682539,"rna_split_half_accuracy":0.9783556547619048,"model_increment":0.006870039682539741,"rna_repeatable_increment":0.47641865079365087,"accuracy_relative_to_rna_repeatability":0.014420173666784787,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.98046875,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":0.48735615079365074,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":0.9996794953323327,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},{"arena":"COMPOSITION_MATCHED_ORIGINAL","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4838020833333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.49311259920634926,"rna_split_half_accuracy":0.9806250000000001,"model_increment":-0.009310515873015957,"rna_repeatable_increment":0.4875124007936508,"accuracy_relative_to_rna_repeatability":-0.019098008292422523,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Context and depth matched"},{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"aligned_example","model_accuracy":0.9909722222222221,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":0.46185763888888887,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":1.0013550645537697,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P1","model_id":"permuted_example","model_accuracy":0.5208506944444444,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.5291145833333333,"rna_split_half_accuracy":0.9903472222222222,"model_increment":-0.008263888888888848,"rna_repeatable_increment":0.4612326388888889,"accuracy_relative_to_rna_repeatability":-0.017916964655399468,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"aligned_example","model_accuracy":0.9817708333333334,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":0.44848958333333333,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":1.0034960960261041,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"},{"arena":"JOINT_STRUCTURE_MATCHED","patient_id":"synthetic_P2","model_id":"permuted_example","model_accuracy":0.4523958333333333,"seeds_evaluable":1.0,"lopo_gene_majority_baseline":0.53328125,"rna_split_half_accuracy":0.9802083333333333,"model_increment":-0.08088541666666677,"rna_repeatable_increment":0.4469270833333333,"accuracy_relative_to_rna_repeatability":-0.18098123761799348,"status":"ESTIMABLE","seeds_requested":1.0,"task":"Within regions"}]},{"name":"run1/question_support.tsv","columns":["patient_id","section_id","seed","arena","questions_selected","questions_judged"],"rows":[{"patient_id":"synthetic_P1","section_id":"synthetic_P1_section","seed":2026080501.0,"arena":"BROAD_STRUCTURE_ALLOWED","questions_selected":957.0,"questions_judged":950.0},{"patient_id":"synthetic_P1","section_id":"synthetic_P1_section","seed":2026080501.0,"arena":"COMPOSITION_MATCHED_ORIGINAL","questions_selected":957.0,"questions_judged":950.0},{"patient_id":"synthetic_P1","section_id":"synthetic_P1_section","seed":2026080501.0,"arena":"JOINT_STRUCTURE_MATCHED","questions_selected":505.0,"questions_judged":500.0},{"patient_id":"synthetic_P2","section_id":"synthetic_P2_section","seed":2026080501.0,"arena":"BROAD_STRUCTURE_ALLOWED","questions_selected":969.0,"questions_judged":967.0},{"patient_id":"synthetic_P2","section_id":"synthetic_P2_section","seed":2026080501.0,"arena":"COMPOSITION_MATCHED_ORIGINAL","questions_selected":969.0,"questions_judged":963.0},{"patient_id":"synthetic_P2","section_id":"synthetic_P2_section","seed":2026080501.0,"arena":"JOINT_STRUCTURE_MATCHED","questions_selected":480.0,"questions_judged":476.0}]}],"notes":["Synthetic software example. No patient data or manuscript evidence.","Patient scores average the saved molecule-split seeds. The normalized increment is not ordinary correlation or all-pairs concordance.","RNA and held-out baseline metrics describe references, not competing prediction models.","This bundle exports saved results without rerunning, clipping, averaging or recalibrating the scientific scores."]}
