更新 math Q1 Q2 实验与特征

This commit is contained in:
2026-09-25 08:58:10 +08:00
parent ca5be95757
commit 83ec3d1a83
168 changed files with 150303 additions and 2874 deletions
@@ -0,0 +1,78 @@
model,baseline,metric,point_delta,bootstrap_median_delta,ci_2_5,ci_97_5,replicates,unit
C1,C0,accuracy,0.015109890109890056,0.014540654071697634,-0.02072183396214814,0.0508771724181045,1000,paired source-video group resample
C1,C0,macro_f1,-0.004691729310120518,-0.005934296634649006,-0.04397414179686003,0.03565646614122406,1000,paired source-video group resample
C1,C0,mae,-0.0554240345954895,-0.05444282293319702,-0.09375483095645905,-0.01503600478172303,1000,paired source-video group resample
C1,C0,rmse,-0.014935817346632296,-0.013258818758929702,-0.06884780543374597,0.046824820732205875,1000,paired source-video group resample
C1,C0,pearson,0.05875682830810547,0.05802242457866669,0.002364498376846314,0.1096881166100502,1000,paired source-video group resample
C1,C0,interval_90_coverage,-0.013736263736263687,-0.01407375396139443,-0.04065771954245597,0.011025889100001162,1000,paired source-video group resample
C1,C0,interval_90_mean_width,-0.6801879405975342,-0.6815143823623657,-0.7549685001373291,-0.5937675774097443,1000,paired source-video group resample
C2,C0,accuracy,0.005494505494505475,0.005369214581201265,-0.02994376576360175,0.0391940871501641,1000,paired source-video group resample
C2,C0,macro_f1,-0.07956378757630245,-0.0790228080874627,-0.11503642522479199,-0.04535533337167957,1000,paired source-video group resample
C2,C0,mae,-0.0174868106842041,-0.01701989769935608,-0.05516543835401535,0.02024096548557281,1000,paired source-video group resample
C2,C0,rmse,0.01013764645327897,0.009857838331185975,-0.04068845414232331,0.06170280906214439,1000,paired source-video group resample
C2,C0,pearson,0.06041759252548218,0.05996474623680115,0.012590093165636064,0.10714993029832838,1000,paired source-video group resample
C2,C0,interval_90_coverage,-0.01098901098901095,-0.010906617201527025,-0.036017915080268074,0.012361034956431066,1000,paired source-video group resample
C2,C0,interval_90_mean_width,-0.5603723526000977,-0.5623085498809814,-0.6364005088806153,-0.4799393117427826,1000,paired source-video group resample
C3,C0,accuracy,0.016483516483516425,0.016563344218070708,-0.01763847840925233,0.050214192879048734,1000,paired source-video group resample
C3,C0,macro_f1,-0.058533700572424185,-0.058234548423185234,-0.09561566579622569,-0.02038892280273245,1000,paired source-video group resample
C3,C0,mae,-0.016335368156433105,-0.015149623155593872,-0.055715514719486235,0.025416558980941768,1000,paired source-video group resample
C3,C0,rmse,0.014782920551098089,0.015419141692077643,-0.038076123069670856,0.07340071085319849,1000,paired source-video group resample
C3,C0,pearson,0.07739311456680298,0.07681527733802795,0.024015167355537416,0.12345541119575498,1000,paired source-video group resample
C3,C0,interval_90_coverage,-0.0027472527472527375,-0.00410961988830838,-0.027750467750793536,0.020954891616959875,1000,paired source-video group resample
C3,C0,interval_90_mean_width,-0.640167236328125,-0.6405612230300903,-0.7109949052333832,-0.5546378016471862,1000,paired source-video group resample
C4,C0,accuracy,0.016483516483516425,0.016415899393667166,-0.017356716818758812,0.04911759336909525,1000,paired source-video group resample
C4,C0,macro_f1,-0.0008760072279735409,-0.0006209418412993495,-0.03810678311317003,0.036679564307315324,1000,paired source-video group resample
C4,C0,mae,-0.04160189628601074,-0.04116714000701904,-0.07862299531698227,-0.0037129849195480426,1000,paired source-video group resample
C4,C0,rmse,0.014384588646868646,0.01554152399820774,-0.041543071226566815,0.08196211635437033,1000,paired source-video group resample
C4,C0,pearson,0.04924803972244263,0.04813617467880249,-0.013162557780742644,0.10131643265485762,1000,paired source-video group resample
C4,C0,interval_90_coverage,-0.01098901098901095,-0.01134751773049647,-0.03755502866983134,0.01479151957106922,1000,paired source-video group resample
C4,C0,interval_90_mean_width,-0.6055305004119873,-0.606930136680603,-0.6910651624202728,-0.5102817654609682,1000,paired source-video group resample
C5,C0,accuracy,0.02335164835164838,0.0231607629427793,-0.007834704176513621,0.05665799312178082,1000,paired source-video group resample
C5,C0,macro_f1,-0.011379078722261737,-0.011081368136203418,-0.05078514978285897,0.03206592123232569,1000,paired source-video group resample
C5,C0,mae,-0.07105213403701782,-0.07029640674591064,-0.1101970300078392,-0.033040848374366764,1000,paired source-video group resample
C5,C0,rmse,-0.043829951679029255,-0.04314210307825145,-0.09518903687579511,0.008231764878183861,1000,paired source-video group resample
C5,C0,pearson,0.09820222854614258,0.09711459279060364,0.054920969903469084,0.14132471680641173,1000,paired source-video group resample
C5,C0,interval_90_coverage,-0.016483516483516425,-0.016346749476701616,-0.041550262655205374,0.006870169627622198,1000,paired source-video group resample
C5,C0,interval_90_mean_width,-0.7618138790130615,-0.7637132406234741,-0.8362245619297027,-0.6784132480621338,1000,paired source-video group resample
C6,C0,accuracy,0.03434065934065933,0.03507210334215988,0.0013193504736129726,0.06677418151987477,1000,paired source-video group resample
C6,C0,macro_f1,0.02417681475733191,0.0245627409794269,-0.012247590393776613,0.0609586676069792,1000,paired source-video group resample
C6,C0,mae,-0.08769041299819946,-0.08800148963928223,-0.12380164712667466,-0.051559735834598545,1000,paired source-video group resample
C6,C0,rmse,-0.0557666607442967,-0.05718944521581815,-0.10411016254737489,-0.011209869721099662,1000,paired source-video group resample
C6,C0,pearson,0.0964621901512146,0.09742751717567444,0.0543783888220787,0.1401650607585907,1000,paired source-video group resample
C6,C0,interval_90_coverage,-0.02335164835164838,-0.024096397645309686,-0.04845282656595448,0.0,1000,paired source-video group resample
C6,C0,interval_90_mean_width,-0.8398799896240234,-0.8415627479553223,-0.8992792367935181,-0.7717564046382904,1000,paired source-video group resample
C6_no_distance,C0,accuracy,0.012362637362637319,0.011665591419527832,-0.021922349201543952,0.0445976159085915,1000,paired source-video group resample
C6_no_distance,C0,macro_f1,-0.041443005147987155,-0.042715995881111246,-0.0793164931295086,-0.0021521221408520616,1000,paired source-video group resample
C6_no_distance,C0,mae,-0.05394166707992554,-0.05345669388771057,-0.08995662033557891,-0.017440022528171555,1000,paired source-video group resample
C6_no_distance,C0,rmse,-0.027334657615164293,-0.027400945882898098,-0.07395318523759227,0.020861843239490783,1000,paired source-video group resample
C6_no_distance,C0,pearson,0.08535701036453247,0.08517509698867798,0.04233360141515732,0.12865879982709885,1000,paired source-video group resample
C6_no_distance,C0,interval_90_coverage,-0.02335164835164838,-0.023421019727422,-0.04858552631578949,-0.0012778505897772484,1000,paired source-video group resample
C6_no_distance,C0,interval_90_mean_width,-0.845958948135376,-0.8460291624069214,-0.8988584756851197,-0.7863011598587036,1000,paired source-video group resample
C6_no_reconstruction,C0,accuracy,0.008241758241758212,0.007412935104671281,-0.02508586214016911,0.041334712950600795,1000,paired source-video group resample
C6_no_reconstruction,C0,macro_f1,-0.038098823880311605,-0.0380264210091098,-0.07862970799872031,0.0008047696861085368,1000,paired source-video group resample
C6_no_reconstruction,C0,mae,-0.03766679763793945,-0.037051618099212646,-0.07584996074438095,0.001131595671176908,1000,paired source-video group resample
C6_no_reconstruction,C0,rmse,-0.004467192313314605,-0.004141038354871618,-0.06167746651854182,0.05485222799945723,1000,paired source-video group resample
C6_no_reconstruction,C0,pearson,0.06950116157531738,0.06919342279434204,0.01576427072286606,0.12347625344991682,1000,paired source-video group resample
C6_no_reconstruction,C0,interval_90_coverage,-0.016483516483516425,-0.01702685733498932,-0.044414535666218,0.008291378102871401,1000,paired source-video group resample
C6_no_reconstruction,C0,interval_90_mean_width,-0.7536470890045166,-0.7535206079483032,-0.8238759577274323,-0.6792047619819641,1000,paired source-video group resample
C6_pointmask,C0,accuracy,0.01098901098901095,0.010928982145223365,-0.021508282476024394,0.04276060564095464,1000,paired source-video group resample
C6_pointmask,C0,macro_f1,-0.004597019933230584,-0.004950783837691097,-0.04252592927395962,0.03481194374641572,1000,paired source-video group resample
C6_pointmask,C0,mae,-0.03887361288070679,-0.03850683569908142,-0.07904898971319199,0.0024588942527770732,1000,paired source-video group resample
C6_pointmask,C0,rmse,0.018223692242232525,0.01828771010893937,-0.03897288944724318,0.07443797524234813,1000,paired source-video group resample
C6_pointmask,C0,pearson,0.04544329643249512,0.04554110765457153,-0.009580320119857787,0.09693647027015685,1000,paired source-video group resample
C6_pointmask,C0,interval_90_coverage,-0.009615384615384581,-0.010281847013516643,-0.034839457133399594,0.015473493244019835,1000,paired source-video group resample
C6_pointmask,C0,interval_90_mean_width,-0.6530871391296387,-0.6545566320419312,-0.7273817896842957,-0.5680322289466858,1000,paired source-video group resample
C7_distill,C0,accuracy,-0.0013736263736263687,-0.0025957181599204904,-0.03561744768641323,0.03287716294160066,1000,paired source-video group resample
C7_distill,C0,macro_f1,-0.03154995832206664,-0.031907844496559035,-0.07075084408889719,0.01119856559020494,1000,paired source-video group resample
C7_distill,C0,mae,-0.0591282844543457,-0.05798792839050293,-0.09516946226358414,-0.022455814480781554,1000,paired source-video group resample
C7_distill,C0,rmse,-0.039041818430481046,-0.03907461260298151,-0.08993284952634278,0.01886223576934703,1000,paired source-video group resample
C7_distill,C0,pearson,0.08209490776062012,0.081440269947052,0.029983995854854586,0.12790912091732023,1000,paired source-video group resample
C7_distill,C0,interval_90_coverage,-0.04258241758241754,-0.0430177622632516,-0.069833904006728,-0.016924193577931214,1000,paired source-video group resample
C7_distill,C0,interval_90_mean_width,-0.8419189453125,-0.8431320190429688,-0.9057650566101074,-0.7694313526153564,1000,paired source-video group resample
C7_group,C0,accuracy,0.016483516483516425,0.01578569810465902,-0.017397203855622368,0.04950699615416091,1000,paired source-video group resample
C7_group,C0,macro_f1,-0.0442503179504562,-0.045386474713565494,-0.08299241213109994,-0.0069366953513426265,1000,paired source-video group resample
C7_group,C0,mae,-0.04738658666610718,-0.046143680810928345,-0.08319220691919327,-0.007835713028907778,1000,paired source-video group resample
C7_group,C0,rmse,-0.011885811847952454,-0.010572588059510002,-0.06285378747754637,0.04485988704499218,1000,paired source-video group resample
C7_group,C0,pearson,0.07194912433624268,0.06962910294532776,0.025362481176853184,0.11925912424921989,1000,paired source-video group resample
C7_group,C0,interval_90_coverage,-0.016483516483516425,-0.01619944693058395,-0.04283849684452092,0.007804650605708006,1000,paired source-video group resample
C7_group,C0,interval_90_mean_width,-0.6982431411743164,-0.6987943649291992,-0.765870851278305,-0.6182706952095032,1000,paired source-video group resample
1 model baseline metric point_delta bootstrap_median_delta ci_2_5 ci_97_5 replicates unit
2 C1 C0 accuracy 0.015109890109890056 0.014540654071697634 -0.02072183396214814 0.0508771724181045 1000 paired source-video group resample
3 C1 C0 macro_f1 -0.004691729310120518 -0.005934296634649006 -0.04397414179686003 0.03565646614122406 1000 paired source-video group resample
4 C1 C0 mae -0.0554240345954895 -0.05444282293319702 -0.09375483095645905 -0.01503600478172303 1000 paired source-video group resample
5 C1 C0 rmse -0.014935817346632296 -0.013258818758929702 -0.06884780543374597 0.046824820732205875 1000 paired source-video group resample
6 C1 C0 pearson 0.05875682830810547 0.05802242457866669 0.002364498376846314 0.1096881166100502 1000 paired source-video group resample
7 C1 C0 interval_90_coverage -0.013736263736263687 -0.01407375396139443 -0.04065771954245597 0.011025889100001162 1000 paired source-video group resample
8 C1 C0 interval_90_mean_width -0.6801879405975342 -0.6815143823623657 -0.7549685001373291 -0.5937675774097443 1000 paired source-video group resample
9 C2 C0 accuracy 0.005494505494505475 0.005369214581201265 -0.02994376576360175 0.0391940871501641 1000 paired source-video group resample
10 C2 C0 macro_f1 -0.07956378757630245 -0.0790228080874627 -0.11503642522479199 -0.04535533337167957 1000 paired source-video group resample
11 C2 C0 mae -0.0174868106842041 -0.01701989769935608 -0.05516543835401535 0.02024096548557281 1000 paired source-video group resample
12 C2 C0 rmse 0.01013764645327897 0.009857838331185975 -0.04068845414232331 0.06170280906214439 1000 paired source-video group resample
13 C2 C0 pearson 0.06041759252548218 0.05996474623680115 0.012590093165636064 0.10714993029832838 1000 paired source-video group resample
14 C2 C0 interval_90_coverage -0.01098901098901095 -0.010906617201527025 -0.036017915080268074 0.012361034956431066 1000 paired source-video group resample
15 C2 C0 interval_90_mean_width -0.5603723526000977 -0.5623085498809814 -0.6364005088806153 -0.4799393117427826 1000 paired source-video group resample
16 C3 C0 accuracy 0.016483516483516425 0.016563344218070708 -0.01763847840925233 0.050214192879048734 1000 paired source-video group resample
17 C3 C0 macro_f1 -0.058533700572424185 -0.058234548423185234 -0.09561566579622569 -0.02038892280273245 1000 paired source-video group resample
18 C3 C0 mae -0.016335368156433105 -0.015149623155593872 -0.055715514719486235 0.025416558980941768 1000 paired source-video group resample
19 C3 C0 rmse 0.014782920551098089 0.015419141692077643 -0.038076123069670856 0.07340071085319849 1000 paired source-video group resample
20 C3 C0 pearson 0.07739311456680298 0.07681527733802795 0.024015167355537416 0.12345541119575498 1000 paired source-video group resample
21 C3 C0 interval_90_coverage -0.0027472527472527375 -0.00410961988830838 -0.027750467750793536 0.020954891616959875 1000 paired source-video group resample
22 C3 C0 interval_90_mean_width -0.640167236328125 -0.6405612230300903 -0.7109949052333832 -0.5546378016471862 1000 paired source-video group resample
23 C4 C0 accuracy 0.016483516483516425 0.016415899393667166 -0.017356716818758812 0.04911759336909525 1000 paired source-video group resample
24 C4 C0 macro_f1 -0.0008760072279735409 -0.0006209418412993495 -0.03810678311317003 0.036679564307315324 1000 paired source-video group resample
25 C4 C0 mae -0.04160189628601074 -0.04116714000701904 -0.07862299531698227 -0.0037129849195480426 1000 paired source-video group resample
26 C4 C0 rmse 0.014384588646868646 0.01554152399820774 -0.041543071226566815 0.08196211635437033 1000 paired source-video group resample
27 C4 C0 pearson 0.04924803972244263 0.04813617467880249 -0.013162557780742644 0.10131643265485762 1000 paired source-video group resample
28 C4 C0 interval_90_coverage -0.01098901098901095 -0.01134751773049647 -0.03755502866983134 0.01479151957106922 1000 paired source-video group resample
29 C4 C0 interval_90_mean_width -0.6055305004119873 -0.606930136680603 -0.6910651624202728 -0.5102817654609682 1000 paired source-video group resample
30 C5 C0 accuracy 0.02335164835164838 0.0231607629427793 -0.007834704176513621 0.05665799312178082 1000 paired source-video group resample
31 C5 C0 macro_f1 -0.011379078722261737 -0.011081368136203418 -0.05078514978285897 0.03206592123232569 1000 paired source-video group resample
32 C5 C0 mae -0.07105213403701782 -0.07029640674591064 -0.1101970300078392 -0.033040848374366764 1000 paired source-video group resample
33 C5 C0 rmse -0.043829951679029255 -0.04314210307825145 -0.09518903687579511 0.008231764878183861 1000 paired source-video group resample
34 C5 C0 pearson 0.09820222854614258 0.09711459279060364 0.054920969903469084 0.14132471680641173 1000 paired source-video group resample
35 C5 C0 interval_90_coverage -0.016483516483516425 -0.016346749476701616 -0.041550262655205374 0.006870169627622198 1000 paired source-video group resample
36 C5 C0 interval_90_mean_width -0.7618138790130615 -0.7637132406234741 -0.8362245619297027 -0.6784132480621338 1000 paired source-video group resample
37 C6 C0 accuracy 0.03434065934065933 0.03507210334215988 0.0013193504736129726 0.06677418151987477 1000 paired source-video group resample
38 C6 C0 macro_f1 0.02417681475733191 0.0245627409794269 -0.012247590393776613 0.0609586676069792 1000 paired source-video group resample
39 C6 C0 mae -0.08769041299819946 -0.08800148963928223 -0.12380164712667466 -0.051559735834598545 1000 paired source-video group resample
40 C6 C0 rmse -0.0557666607442967 -0.05718944521581815 -0.10411016254737489 -0.011209869721099662 1000 paired source-video group resample
41 C6 C0 pearson 0.0964621901512146 0.09742751717567444 0.0543783888220787 0.1401650607585907 1000 paired source-video group resample
42 C6 C0 interval_90_coverage -0.02335164835164838 -0.024096397645309686 -0.04845282656595448 0.0 1000 paired source-video group resample
43 C6 C0 interval_90_mean_width -0.8398799896240234 -0.8415627479553223 -0.8992792367935181 -0.7717564046382904 1000 paired source-video group resample
44 C6_no_distance C0 accuracy 0.012362637362637319 0.011665591419527832 -0.021922349201543952 0.0445976159085915 1000 paired source-video group resample
45 C6_no_distance C0 macro_f1 -0.041443005147987155 -0.042715995881111246 -0.0793164931295086 -0.0021521221408520616 1000 paired source-video group resample
46 C6_no_distance C0 mae -0.05394166707992554 -0.05345669388771057 -0.08995662033557891 -0.017440022528171555 1000 paired source-video group resample
47 C6_no_distance C0 rmse -0.027334657615164293 -0.027400945882898098 -0.07395318523759227 0.020861843239490783 1000 paired source-video group resample
48 C6_no_distance C0 pearson 0.08535701036453247 0.08517509698867798 0.04233360141515732 0.12865879982709885 1000 paired source-video group resample
49 C6_no_distance C0 interval_90_coverage -0.02335164835164838 -0.023421019727422 -0.04858552631578949 -0.0012778505897772484 1000 paired source-video group resample
50 C6_no_distance C0 interval_90_mean_width -0.845958948135376 -0.8460291624069214 -0.8988584756851197 -0.7863011598587036 1000 paired source-video group resample
51 C6_no_reconstruction C0 accuracy 0.008241758241758212 0.007412935104671281 -0.02508586214016911 0.041334712950600795 1000 paired source-video group resample
52 C6_no_reconstruction C0 macro_f1 -0.038098823880311605 -0.0380264210091098 -0.07862970799872031 0.0008047696861085368 1000 paired source-video group resample
53 C6_no_reconstruction C0 mae -0.03766679763793945 -0.037051618099212646 -0.07584996074438095 0.001131595671176908 1000 paired source-video group resample
54 C6_no_reconstruction C0 rmse -0.004467192313314605 -0.004141038354871618 -0.06167746651854182 0.05485222799945723 1000 paired source-video group resample
55 C6_no_reconstruction C0 pearson 0.06950116157531738 0.06919342279434204 0.01576427072286606 0.12347625344991682 1000 paired source-video group resample
56 C6_no_reconstruction C0 interval_90_coverage -0.016483516483516425 -0.01702685733498932 -0.044414535666218 0.008291378102871401 1000 paired source-video group resample
57 C6_no_reconstruction C0 interval_90_mean_width -0.7536470890045166 -0.7535206079483032 -0.8238759577274323 -0.6792047619819641 1000 paired source-video group resample
58 C6_pointmask C0 accuracy 0.01098901098901095 0.010928982145223365 -0.021508282476024394 0.04276060564095464 1000 paired source-video group resample
59 C6_pointmask C0 macro_f1 -0.004597019933230584 -0.004950783837691097 -0.04252592927395962 0.03481194374641572 1000 paired source-video group resample
60 C6_pointmask C0 mae -0.03887361288070679 -0.03850683569908142 -0.07904898971319199 0.0024588942527770732 1000 paired source-video group resample
61 C6_pointmask C0 rmse 0.018223692242232525 0.01828771010893937 -0.03897288944724318 0.07443797524234813 1000 paired source-video group resample
62 C6_pointmask C0 pearson 0.04544329643249512 0.04554110765457153 -0.009580320119857787 0.09693647027015685 1000 paired source-video group resample
63 C6_pointmask C0 interval_90_coverage -0.009615384615384581 -0.010281847013516643 -0.034839457133399594 0.015473493244019835 1000 paired source-video group resample
64 C6_pointmask C0 interval_90_mean_width -0.6530871391296387 -0.6545566320419312 -0.7273817896842957 -0.5680322289466858 1000 paired source-video group resample
65 C7_distill C0 accuracy -0.0013736263736263687 -0.0025957181599204904 -0.03561744768641323 0.03287716294160066 1000 paired source-video group resample
66 C7_distill C0 macro_f1 -0.03154995832206664 -0.031907844496559035 -0.07075084408889719 0.01119856559020494 1000 paired source-video group resample
67 C7_distill C0 mae -0.0591282844543457 -0.05798792839050293 -0.09516946226358414 -0.022455814480781554 1000 paired source-video group resample
68 C7_distill C0 rmse -0.039041818430481046 -0.03907461260298151 -0.08993284952634278 0.01886223576934703 1000 paired source-video group resample
69 C7_distill C0 pearson 0.08209490776062012 0.081440269947052 0.029983995854854586 0.12790912091732023 1000 paired source-video group resample
70 C7_distill C0 interval_90_coverage -0.04258241758241754 -0.0430177622632516 -0.069833904006728 -0.016924193577931214 1000 paired source-video group resample
71 C7_distill C0 interval_90_mean_width -0.8419189453125 -0.8431320190429688 -0.9057650566101074 -0.7694313526153564 1000 paired source-video group resample
72 C7_group C0 accuracy 0.016483516483516425 0.01578569810465902 -0.017397203855622368 0.04950699615416091 1000 paired source-video group resample
73 C7_group C0 macro_f1 -0.0442503179504562 -0.045386474713565494 -0.08299241213109994 -0.0069366953513426265 1000 paired source-video group resample
74 C7_group C0 mae -0.04738658666610718 -0.046143680810928345 -0.08319220691919327 -0.007835713028907778 1000 paired source-video group resample
75 C7_group C0 rmse -0.011885811847952454 -0.010572588059510002 -0.06285378747754637 0.04485988704499218 1000 paired source-video group resample
76 C7_group C0 pearson 0.07194912433624268 0.06962910294532776 0.025362481176853184 0.11925912424921989 1000 paired source-video group resample
77 C7_group C0 interval_90_coverage -0.016483516483516425 -0.01619944693058395 -0.04283849684452092 0.007804650605708006 1000 paired source-video group resample
78 C7_group C0 interval_90_mean_width -0.6982431411743164 -0.6987943649291992 -0.765870851278305 -0.6182706952095032 1000 paired source-video group resample