ForesightLM-Phase1 / outputs /phase1 /bootstrap /phase1_bootstrap_summary.csv
Mandotosh's picture
Upload bootstrap robustness results
7e0e88b verified
Raw
History Blame Contribute Delete
6.91 kB
comparison,metric,model_a,model_b,mean_a,mean_b,delta_a_minus_b,ci_low,ci_high,p_direction,n_prompts,n_bootstrap
v2_minus_core,prompt_sim,v2,core,0.42742575221539786,0.417741510256504,0.009684241958893836,-0.014334722388302907,0.03340032646257895,0.2117,120,10000
v2_minus_core,topic_drift,v2,core,0.5725742477846022,0.582258489743496,-0.009684241958893836,-0.033228457572404294,0.0140339194383705,0.2097,120,10000
v2_minus_core,rep_adj,v2,core,0.010224358974358975,0.024644795657726693,-0.014420436683367717,-0.03675067694518125,0.005540412552804795,0.0807,120,10000
v2_minus_core,rep_pair,v2,core,0.008393365893365893,0.013133639987088263,-0.0047402740937223685,-0.01998703914221155,0.008390627412179132,0.2712,120,10000
v2_minus_core,progression_band,v2,core,0.6360696479446479,0.6674425287356323,-0.03137288079098424,-0.09394879469987227,0.0325977232976155,0.1646,120,10000
v2_minus_core,lexical_loop_penalty,v2,core,0.07701128526080024,0.11459236275146452,-0.037581077490664284,-0.077373644597447,0.0005587195890821763,0.0271,120,10000
v2_minus_core,exact_lexical_loop,v2,core,0.2833333333333333,0.31666666666666665,-0.03333333333333333,-0.14166666666666666,0.075,0.3033,120,10000
v2_minus_core,semantic_loop,v2,core,0.06666666666666667,0.09166666666666666,-0.025,-0.09166666666666666,0.041666666666666664,0.2715,120,10000
v2_minus_core,prompt_drift,v2,core,0.16666666666666666,0.19166666666666668,-0.025,-0.1,0.05,0.2912,120,10000
v2_minus_core,weak_progression,v2,core,0.4,0.35833333333333334,0.041666666666666664,-0.075,0.15833333333333333,0.2627,120,10000
v2_calibrated_c5_minus_core,prompt_sim,v2_calibrated_c5,core,0.4953933523967862,0.417741510256504,0.07765184214028219,0.052679028646477194,0.10289676335446225,0.0,120,10000
v2_calibrated_c5_minus_core,topic_drift,v2_calibrated_c5,core,0.5046066476032138,0.582258489743496,-0.07765184214028219,-0.10261079253124383,-0.05282198344318506,0.0,120,10000
v2_calibrated_c5_minus_core,rep_adj,v2_calibrated_c5,core,0.0,0.024644795657726693,-0.024644795657726693,-0.04381924489144317,-0.009019416507024272,0.0,120,10000
v2_calibrated_c5_minus_core,rep_pair,v2_calibrated_c5,core,0.003630952380952381,0.013133639987088263,-0.00950268760613588,-0.023706889534475736,0.0017172961138478369,0.0538,120,10000
v2_calibrated_c5_minus_core,progression_band,v2_calibrated_c5,core,0.872714947089947,0.6674425287356323,0.20527241835431492,0.14412666769293928,0.26460901010308346,0.0,120,10000
v2_calibrated_c5_minus_core,lexical_loop_penalty,v2_calibrated_c5,core,0.10273307336652525,0.11459236275146452,-0.011859289384939289,-0.06206824574604235,0.03688690511192931,0.325,120,10000
v2_calibrated_c5_minus_core,exact_lexical_loop,v2_calibrated_c5,core,0.325,0.31666666666666665,0.008333333333333333,-0.1,0.11666666666666667,0.4752,120,10000
v2_calibrated_c5_minus_core,semantic_loop,v2_calibrated_c5,core,0.025,0.09166666666666666,-0.06666666666666667,-0.125,-0.008333333333333333,0.0179,120,10000
v2_calibrated_c5_minus_core,prompt_drift,v2_calibrated_c5,core,0.075,0.19166666666666668,-0.11666666666666667,-0.18333333333333332,-0.05,0.0002,120,10000
v2_calibrated_c5_minus_core,weak_progression,v2_calibrated_c5,core,0.06666666666666667,0.35833333333333334,-0.2916666666666667,-0.39166666666666666,-0.19166666666666668,0.0,120,10000
v2_calibrated_c5_minus_baseline,prompt_sim,v2_calibrated_c5,baseline,0.4953933523967862,0.419695561643069,0.07569779075371723,0.04942825103295036,0.10246047055271143,0.0,120,10000
v2_calibrated_c5_minus_baseline,topic_drift,v2_calibrated_c5,baseline,0.5046066476032138,0.580304438356931,-0.07569779075371723,-0.10292517608962953,-0.04924626664413761,0.0,120,10000
v2_calibrated_c5_minus_baseline,rep_adj,v2_calibrated_c5,baseline,0.0,0.023875849190250813,-0.023875849190250813,-0.04554280764029749,-0.006909313725490198,0.0002,120,10000
v2_calibrated_c5_minus_baseline,rep_pair,v2_calibrated_c5,baseline,0.003630952380952381,0.014127715262350151,-0.01049676288139777,-0.025938013471672194,0.0009135992519613183,0.0412,120,10000
v2_calibrated_c5_minus_baseline,progression_band,v2_calibrated_c5,baseline,0.872714947089947,0.667454491854213,0.20526045523573413,0.15239755233990146,0.256650521319639,0.0,120,10000
v2_calibrated_c5_minus_baseline,lexical_loop_penalty,v2_calibrated_c5,baseline,0.10273307336652525,0.11296329904449201,-0.01023022567796677,-0.06751814638009117,0.044096100544658186,0.3658,120,10000
v2_calibrated_c5_minus_baseline,exact_lexical_loop,v2_calibrated_c5,baseline,0.325,0.3,0.025,-0.09166666666666666,0.14166666666666666,0.3551,120,10000
v2_calibrated_c5_minus_baseline,semantic_loop,v2_calibrated_c5,baseline,0.025,0.06666666666666667,-0.041666666666666664,-0.09166666666666666,0.008333333333333333,0.0606,120,10000
v2_calibrated_c5_minus_baseline,prompt_drift,v2_calibrated_c5,baseline,0.075,0.20833333333333334,-0.13333333333333333,-0.2,-0.06666666666666667,0.0,120,10000
v2_calibrated_c5_minus_baseline,weak_progression,v2_calibrated_c5,baseline,0.06666666666666667,0.375,-0.30833333333333335,-0.4,-0.21666666666666667,0.0,120,10000
baseline_rep_penalty_minus_baseline,prompt_sim,baseline_rep_penalty,baseline,0.3558435012275974,0.419695561643069,-0.06385206041547159,-0.09164656913722866,-0.035541312434167305,0.0,120,10000
baseline_rep_penalty_minus_baseline,topic_drift,baseline_rep_penalty,baseline,0.6441564987724026,0.580304438356931,0.06385206041547159,0.035899566583878675,0.09180333533945181,0.0,120,10000
baseline_rep_penalty_minus_baseline,rep_adj,baseline_rep_penalty,baseline,0.0,0.023875849190250813,-0.023875849190250813,-0.04550146495379761,-0.007170868347338936,0.0003,120,10000
baseline_rep_penalty_minus_baseline,rep_pair,baseline_rep_penalty,baseline,0.0,0.014127715262350151,-0.014127715262350151,-0.029040374770597892,-0.003399180457623667,0.0,120,10000
baseline_rep_penalty_minus_baseline,progression_band,baseline_rep_penalty,baseline,0.4197420634920635,0.667454491854213,-0.2477124283621495,-0.32201206574583857,-0.17468596335804434,0.0,120,10000
baseline_rep_penalty_minus_baseline,lexical_loop_penalty,baseline_rep_penalty,baseline,0.0002543139730639731,0.11296329904449201,-0.11270898507142804,-0.16070996851914124,-0.07367638869322533,0.0,120,10000
baseline_rep_penalty_minus_baseline,exact_lexical_loop,baseline_rep_penalty,baseline,0.0,0.3,-0.3,-0.38333333333333336,-0.21666666666666667,0.0,120,10000
baseline_rep_penalty_minus_baseline,semantic_loop,baseline_rep_penalty,baseline,0.0,0.06666666666666667,-0.06666666666666667,-0.11666666666666667,-0.025,0.0003,120,10000
baseline_rep_penalty_minus_baseline,prompt_drift,baseline_rep_penalty,baseline,0.24166666666666667,0.20833333333333334,0.03333333333333333,-0.041666666666666664,0.11666666666666667,0.2279,120,10000
baseline_rep_penalty_minus_baseline,weak_progression,baseline_rep_penalty,baseline,0.7083333333333334,0.375,0.3333333333333333,0.20833333333333334,0.45,0.0,120,10000