{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":6,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":6,"direct_label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline (scores rank; they never assert a category)","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12"},"query_hash":"3b7e32a92e2b","filters":{"venue":"Proceedings - Working Conference on Reverse Engineering"}},"results":[{"id":"W1989130222","doi":"10.1109/wcre.2007.15","title":"Clone Detection via Structural Abstraction","year":2007,"lang":"en","type":"article","venue":"Proceedings - Working Conference on Reverse Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":44,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of British Columbia","funders":"","keywords":"Abstraction; Programming language; Computer science; Compiler; Cloning (programming); clone (Java method); Syntax; Representation (politics); Intermediate language; Code (set theory); Theoretical computer science; Abstract syntax tree; Code generation; Line (geometry); Algorithm; Artificial intelligence; Parsing; Mathematics; Operating system; Biology; Genetics","retraction":null,"screen_n_in":null,"score":{"opus":0.02281679062459205,"gpt":0.2538246328404317,"spread":0.2310078422158396,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006936364,0.0002887682,0.0002024113,0.0005100538,0.000153417,0.0003260191,0.0006145533,0.0001565832,0.00001617931],"category_scores_gemma":[0.00030314,0.0003126945,0.00007201943,0.0008998132,0.00002179254,0.0006319904,0.0001357836,0.0007255538,0.00006219829],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002987118,"about_ca_system_score_gemma":0.0000265051,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002207345,"about_ca_topic_score_gemma":0.00000212241,"domain_scores_codex":[0.9978643,0.000003646706,0.0003079335,0.0005530739,0.0005732787,0.0006977649],"domain_scores_gemma":[0.9990574,0.0002070293,0.0001012073,0.0002432488,0.0001821553,0.0002089317],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001226343,0.00008293217,0.01529103,0.0003730491,0.0001545304,0.000148418,0.002080765,0.009181779,0.554019,0.04037028,0.0001831861,0.3779924],"study_design_scores_gemma":[0.001025898,0.0003499553,0.1456656,0.0007391002,0.00002405616,0.0003472165,0.0001534036,0.7005761,0.1418946,0.0009127151,0.006574987,0.001736355],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3802826,0.0000278157,0.6159376,0.00009367928,0.001480801,0.0002052448,2.194007e-7,0.001254773,0.0007172545],"genre_scores_gemma":[0.9810838,0.0000103438,0.01844559,0.00003212103,0.0002874612,0.00001619646,9.461702e-7,0.00003532098,0.00008827758],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6913943,"threshold_uncertainty_score":0.9999325,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2154473425","doi":"10.1109/wcre.2007.5","title":"A Model to Understand the Building and Running Inter-Dependencies of Software","year":2007,"lang":"en","type":"article","venue":"Proceedings - Working Conference on Reverse Engineering","topic":"Advanced Software Engineering Methodologies","field":"Computer Science","cited_by":37,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Victoria","funders":"","keywords":"Computer science; Dependency (UML); Software; Software visualization; Modular design; Software architecture; Software engineering; Software framework; Software system; Architecture; Visualization; Software construction; Programming language; Data mining","retraction":null,"screen_n_in":null,"score":{"opus":0.09417299663578567,"gpt":0.2972885560122112,"spread":0.2031155593764255,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001244221,0.0002827737,0.0002983753,0.0003314986,0.0001298934,0.0001669694,0.0008006341,0.00009150161,0.000001027443],"category_scores_gemma":[0.001142721,0.0002423819,0.00005370742,0.0006187584,0.00005472435,0.0003501306,0.0004755471,0.0004136394,0.000001142311],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001140647,"about_ca_system_score_gemma":0.0000332443,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000003928846,"about_ca_topic_score_gemma":0.000001010022,"domain_scores_codex":[0.9983646,0.000005409165,0.0003314113,0.0004634846,0.0003270632,0.0005080689],"domain_scores_gemma":[0.9988284,0.0005204833,0.0001264677,0.0002552122,0.0001376601,0.0001318271],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00004929539,0.00002387233,0.001144471,0.0002936033,0.000075081,0.000017283,0.01546531,0.4178361,0.04863739,0.4617076,0.0000742042,0.0546758],"study_design_scores_gemma":[0.0007015332,0.0003135135,0.001807095,0.003690283,0.0000466581,0.0001017605,0.00475751,0.9148586,0.0406728,0.03067971,0.0007416398,0.001628936],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06221088,0.0000888801,0.9363751,0.0002124497,0.0002380806,0.0001911162,5.147542e-7,0.0005007233,0.000182297],"genre_scores_gemma":[0.5401838,0.00001552793,0.4596809,0.00004976844,0.00002532002,0.000005101123,6.760519e-8,0.00001676099,0.00002282082],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.4970224,"threshold_uncertainty_score":0.9884045,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2156491770","doi":"10.1109/wcre.2007.52","title":"Visualizing Software Architecture Evolution Using Change-Sets","year":2007,"lang":"en","type":"article","venue":"Proceedings - Working Conference on Reverse Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":27,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Victoria","funders":"","keywords":"Software evolution; Computer science; Software architecture; Java; Architecture; Software system; Software; Set (abstract data type); Resource-oriented architecture; Software engineering; Multilayered architecture; Software architecture description; Reference architecture; Programming language; Component-based software engineering; Software construction","retraction":null,"screen_n_in":null,"score":{"opus":0.06748153835963974,"gpt":0.2986947521289853,"spread":0.2312132137693456,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001072832,0.0004514675,0.0003373987,0.0009342883,0.0002221264,0.0004041565,0.001065526,0.0002212443,0.00001171637],"category_scores_gemma":[0.0009646612,0.0004919295,0.0001144476,0.001570277,0.00003772019,0.0006450734,0.0003883707,0.0008788668,0.00003263807],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005517076,"about_ca_system_score_gemma":0.00007684439,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003872202,"about_ca_topic_score_gemma":0.00000199435,"domain_scores_codex":[0.9967989,0.000009287473,0.0003893362,0.0008070449,0.000822554,0.001172903],"domain_scores_gemma":[0.9986025,0.0003342822,0.0001370601,0.0003661966,0.0002390114,0.0003209336],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002228067,0.0004916858,0.1396542,0.002360309,0.0004844727,0.0007322768,0.01874802,0.0472368,0.1118793,0.3101065,0.0006804079,0.3674032],"study_design_scores_gemma":[0.001917309,0.0005100453,0.06250197,0.007937794,0.00007280082,0.0005506318,0.000522274,0.8899076,0.01650934,0.003438081,0.01184401,0.004288152],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1374018,0.0001550111,0.859333,0.0001296769,0.0008309873,0.0003691997,9.674188e-7,0.001561447,0.0002178825],"genre_scores_gemma":[0.8266796,0.00001197172,0.1727391,0.00007241691,0.0003747931,0.00002837043,0.000001621459,0.00006365917,0.00002850313],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8426708,"threshold_uncertainty_score":0.9997532,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2089991372","doi":"10.1109/wcre.2007.32","title":"Lossless Comparison of Nested Software Decompositions","year":2007,"lang":"en","type":"article","venue":"Proceedings - Working Conference on Reverse Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"York University","funders":"","keywords":"Computer science; Decomposition; Lossless compression; Software; Cluster analysis; Software system; Theoretical computer science; Nested set model; Algorithm; Data mining; Distributed computing; Programming language; Data compression; Relational database; Artificial intelligence","retraction":null,"screen_n_in":null,"score":{"opus":0.04362021653374086,"gpt":0.3034449051996365,"spread":0.2598246886658956,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006573893,0.0003034148,0.0003988048,0.0006184088,0.0001180073,0.0001980636,0.001112847,0.0001482645,0.00001483046],"category_scores_gemma":[0.0006716773,0.000335237,0.0000996052,0.001321802,0.00004674401,0.0003549554,0.0002585732,0.0005856897,0.00002757438],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001556911,"about_ca_system_score_gemma":0.00006255525,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000009300313,"about_ca_topic_score_gemma":0.000001233101,"domain_scores_codex":[0.9976488,0.000005869397,0.0005089479,0.000511134,0.0006350082,0.000690208],"domain_scores_gemma":[0.9983174,0.0006095754,0.0001608645,0.0003407268,0.0003364706,0.0002349712],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001308824,0.000896175,0.3552408,0.00142561,0.0003403723,0.0001392291,0.009291124,0.03249528,0.09867981,0.3976923,0.001742861,0.1019256],"study_design_scores_gemma":[0.002447808,0.0008813132,0.1728998,0.00674642,0.00008504334,0.0001613545,0.0009804892,0.6079419,0.1942524,0.001179314,0.009070804,0.003353288],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2633045,0.00008256082,0.7339882,0.0001313469,0.0005334413,0.0002557207,0.000001368153,0.000991444,0.0007113659],"genre_scores_gemma":[0.8955424,0.000008347134,0.1042423,0.00002113911,0.00008823669,0.00001688771,0.000002359004,0.00003437298,0.00004392797],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6322379,"threshold_uncertainty_score":0.9999099,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2007655484","doi":"10.1109/wcre.2007.10","title":"An Amalgamated Dynamic and Static Architecture Reconstruction Framework to Control Component Interactions 259","year":2007,"lang":"en","type":"article","venue":"Proceedings - Working Conference on Reverse Engineering","topic":"Advanced Software Engineering Methodologies","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McMaster University","funders":"","keywords":"Computer science; Profiling (computer programming); Static analysis; Software system; Component-based software engineering; Software; Software architecture; Software engineering; Component (thermodynamics); Distributed computing; Programming language","retraction":null,"screen_n_in":null,"score":{"opus":0.02255567418482396,"gpt":0.2908430627538822,"spread":0.2682873885690582,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006379655,0.0003331969,0.0003280461,0.0005197301,0.0001379681,0.0002798641,0.0004768411,0.0001152126,0.000005224529],"category_scores_gemma":[0.0006613053,0.0003502284,0.00004306853,0.000617402,0.00003613104,0.000452891,0.00009687542,0.0007268528,0.000005794422],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001804181,"about_ca_system_score_gemma":0.00002010826,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004441668,"about_ca_topic_score_gemma":0.000001984078,"domain_scores_codex":[0.9982482,0.00001782375,0.0003487373,0.0006043192,0.0002401634,0.0005407683],"domain_scores_gemma":[0.9986438,0.0005746085,0.0001228032,0.0002685878,0.0001178516,0.0002723674],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001650499,0.00008131449,0.0007557797,0.0002080443,0.0001184955,0.00004132609,0.005667752,0.3698558,0.0906691,0.0955634,0.00001916563,0.4368548],"study_design_scores_gemma":[0.0008279156,0.0004993439,0.007268168,0.002954253,0.00004499356,0.0003695787,0.0008688999,0.9624305,0.005890764,0.01596276,0.001330483,0.001552382],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1281793,0.0000215006,0.8691889,0.0003523184,0.0008174916,0.000285793,0.000001550589,0.001089495,0.00006370564],"genre_scores_gemma":[0.5476748,0.00001032538,0.4521336,0.00009059111,0.00004395902,0.00001827887,7.710527e-7,0.00002055975,0.000007046071],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5925747,"threshold_uncertainty_score":0.999895,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1975143584","doi":"10.1109/wcre.2007.27","title":"Examining the Effects of Global Data Usage on Software Maintainability","year":2007,"lang":"en","type":"article","venue":"Proceedings - Working Conference on Reverse Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Waterloo","funders":"","keywords":"Maintainability; Computer science; Software maintenance; Source code; Software; Software evolution; Variable (mathematics); Source lines of code; Task (project management); Software metric; Software development; Software engineering; Software construction; Operating system; Systems engineering; Engineering","retraction":null,"screen_n_in":null,"score":{"opus":0.04660345742979925,"gpt":0.2815427781236253,"spread":0.2349393206938261,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.002262407,0.0003468856,0.000321149,0.0001651755,0.0001391472,0.000240222,0.003092343,0.0001316027,0.000005002168],"category_scores_gemma":[0.007551095,0.0002931495,0.00005913149,0.001255477,0.00007909133,0.0003986871,0.001069389,0.0006622117,0.00001413952],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003399869,"about_ca_system_score_gemma":0.00009829232,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000180378,"about_ca_topic_score_gemma":0.000001458522,"domain_scores_codex":[0.9971026,0.00001673665,0.0003846339,0.000850861,0.0008372596,0.0008078844],"domain_scores_gemma":[0.9962676,0.002040996,0.0001443627,0.001165946,0.0001900661,0.0001909847],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.0001763259,0.0004485941,0.3602753,0.002977516,0.0003331305,0.000324814,0.002975781,0.004690546,0.007706849,0.3920078,0.001664259,0.2264191],"study_design_scores_gemma":[0.002376918,0.001743359,0.7430359,0.0087215,0.00009681247,0.000131939,0.000672618,0.2128767,0.01763795,0.002702043,0.007289848,0.002714455],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4871111,0.0001018947,0.509217,0.0002010864,0.001069809,0.0006524768,0.000004158488,0.0009176753,0.0007249054],"genre_scores_gemma":[0.9717863,0.00001467917,0.02792007,0.00006756256,0.0001335639,0.00002064521,0.000002020326,0.00002739324,0.00002781378],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4846752,"threshold_uncertainty_score":0.9999521,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}