{"meta":{"query_hash":"fed833773f56","filters":{"venue":"SuperIntelligence - Robotics - Safety & Alignment"},"cohort_total":14,"direct_labels_cover":0,"predictions_cover":14,"exported":14,"export_cap":100000,"truncated":false,"label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12"},"permalink":"https://metacan.xera.ac/q/fed833773f56","api":"https://metacan.xera.ac/api/v1/cohort?venue=SuperIntelligence+-+Robotics+-+Safety+%26+Alignment"},"results":[{"id":"W4408666990","doi":"10.70777/si.v2i1.13975","title":"Effective Mitigations for Systemic Risks from General-Purpose AI","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Risk and Safety Analysis","field":"Decision Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Institute on Governance","funders":"","keywords":"Systemic risk; Computer science; Economics","score_opus":0.07041623951217256,"score_gpt":0.4075346826216562,"score_spread":0.33711844310948363,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4408666990","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.017567614,0.0014078635,0.96878785,0.007796708,0.0011681728,0.0017996938,0.00031648375,0.00009325221,0.001062346],"genre_scores_gemma":[0.97084904,0.0015447537,0.020777833,0.0012134666,0.0002470868,0.0004029308,0.00013256799,0.00003615212,0.004796151],"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.99426275,0.00052477553,0.0018342055,0.0013362138,0.001362653,0.0006793958],"domain_scores_gemma":[0.9936231,0.0034723287,0.0002812947,0.0015895211,0.00075940415,0.00027437444],"candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.002189176,0.0004730692,0.00093872304,0.00045412354,0.0007807153,0.00037668282,0.0014632635,0.00025213123,0.000416862],"category_scores_gemma":[0.0011429194,0.00039552152,0.00079122663,0.0015062983,0.00027392845,0.00033774864,0.000290244,0.000283523,0.0005695809],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00025950882,0.00026549314,0.010376178,0.00003473516,0.000652473,0.000017870907,0.0014152052,0.82210267,0.0038372567,0.059348036,0.0036473337,0.09804324],"study_design_scores_gemma":[0.0015047146,0.0004305123,0.011433134,0.0005176209,0.0012071761,0.000024544273,0.0074947667,0.2821409,0.0619232,0.5525585,0.0791002,0.0016647585],"about_ca_topic_score_codex":0.0008051616,"about_ca_topic_score_gemma":0.00021296412,"teacher_disagreement_score":0.95328146,"about_ca_system_score_codex":0.00047807585,"about_ca_system_score_gemma":0.00023995091,"threshold_uncertainty_score":0.9998497},"labels":[],"label_agreement":null},{"id":"W4408666991","doi":"10.70777/si.v2i1.13799","title":"Can a Bayesian Oracle Prevent Harm from an Agent?","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Blockchain Technology Applications and Security","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Université de Montréal","funders":"","keywords":"Harm; Oracle; Bayesian probability; Computer science; Psychology; Artificial intelligence; Social psychology","score_opus":0.01427476112546157,"score_gpt":0.2769327674598674,"score_spread":0.26265800633440584,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4408666991","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.006838423,0.00041462143,0.9758202,0.014547921,0.00049650314,0.00061863655,0.00003881851,0.00043166644,0.0007932006],"genre_scores_gemma":[0.9023917,0.0004094417,0.09566878,0.0010043489,0.00004261609,0.000118533004,0.000029862837,0.000017525394,0.00031722177],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9973372,0.000113639304,0.00061921036,0.0010000758,0.0003589666,0.00057090085],"domain_scores_gemma":[0.9974709,0.000089790294,0.00009632735,0.0020317438,0.000094179195,0.00021704043],"candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0003597389,0.0003284859,0.00033660515,0.00015943006,0.00040529144,0.00014899787,0.0021983278,0.00021463554,0.00012098009],"category_scores_gemma":[0.000022983171,0.00033885147,0.00012771838,0.0007203518,0.00018089998,0.0002000797,0.00075156544,0.00030337492,0.00007463865],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000015003906,0.0009720396,0.00089071813,0.000020759899,0.00014170623,0.000027466536,0.0024258725,0.013378172,0.001320124,0.84421587,0.00034403874,0.13624825],"study_design_scores_gemma":[0.00054067414,0.00040218214,0.004691902,0.00017913146,0.00012047174,0.000019333113,0.0013535152,0.38244927,0.106270745,0.44756857,0.054998957,0.0014052491],"about_ca_topic_score_codex":0.00044802023,"about_ca_topic_score_gemma":0.00032260735,"teacher_disagreement_score":0.89555323,"about_ca_system_score_codex":0.00030016468,"about_ca_system_score_gemma":0.00018616198,"threshold_uncertainty_score":0.99990636},"labels":[],"label_agreement":null},{"id":"W4410907699","doi":"10.70777/si.v2i2.14611","title":"Pitfalls of Evidence-Based AI Policy","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"","funders":"","keywords":"Computer science; Political science","score_opus":0.23126267499608177,"score_gpt":0.4610154853350626,"score_spread":0.22975281033898082,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4410907699","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0740966,0.004155743,0.7119298,0.19706364,0.0033036296,0.0034665505,0.00003918934,0.00027584872,0.0056690234],"genre_scores_gemma":[0.9847952,0.0023596513,0.005429039,0.005394419,0.0002679825,0.000049267648,0.000025503323,0.000027862243,0.0016510442],"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.99690366,0.00011475168,0.0012814149,0.00052173564,0.00059483934,0.00058357883],"domain_scores_gemma":[0.99758804,0.0005362899,0.00014656286,0.0008845735,0.0005705809,0.00027394257],"candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.000685824,0.00030368872,0.0005516469,0.0004509739,0.00017379482,0.00003012771,0.00032398425,0.00020913199,0.00042277234],"category_scores_gemma":[0.0008677357,0.00025752492,0.00025929027,0.0010624888,0.00031263684,0.00017092416,0.00008980784,0.00031115304,0.00014879732],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0027697284,0.0030710008,0.142987,0.0035513893,0.000541561,0.000054843844,0.005191305,0.18815172,0.02723128,0.19144732,0.0033674645,0.4316354],"study_design_scores_gemma":[0.00022997815,0.0016873433,0.010390313,0.005697353,0.0004034737,0.000024299281,0.0043932525,0.01909823,0.91968757,0.015989088,0.021634124,0.0007649619],"about_ca_topic_score_codex":0.0018860811,"about_ca_topic_score_gemma":0.000057010337,"teacher_disagreement_score":0.91069865,"about_ca_system_score_codex":0.00073547295,"about_ca_system_score_gemma":0.0023060988,"threshold_uncertainty_score":0.9999877},"labels":[],"label_agreement":null},{"id":"W4411012657","doi":"10.70777/si.v2i2.14755","title":"The First International AI Safety Report","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":27,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Université de Montréal","funders":"Mozilla Foundation; Royal Society; DeepMind","keywords":"Forensic engineering; Engineering","score_opus":0.022935767484034593,"score_gpt":0.361102392748586,"score_spread":0.33816662526455143,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4411012657","genre_codex":"commentary","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.00054120435,0.0005409989,0.07639238,0.5307176,0.008911761,0.0009013967,0.000029362825,0.00019430176,0.38177097],"genre_scores_gemma":[0.9360189,0.016243158,0.0014776223,0.0074972822,0.00083390786,0.000041471525,0.00003436574,0.00003597739,0.037817333],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","domain_scores_codex":[0.996843,0.00015880601,0.00079642,0.00043977358,0.0010760056,0.00068603095],"domain_scores_gemma":[0.99783206,0.0007078899,0.000141704,0.0005664433,0.0005487502,0.00020316787],"candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.0029755929,0.00023074112,0.0002483101,0.00006971537,0.003337622,0.0005622286,0.0011714563,0.0002358385,0.00031170747],"category_scores_gemma":[0.001470247,0.0001943494,0.00022441974,0.00040545253,0.0009558807,0.0003383554,0.0003171214,0.00046553355,0.00009485552],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0000716083,0.00011713997,0.0040819934,0.0000110872925,0.00019018444,0.00005158601,0.007026908,0.0071741887,0.000013785402,0.966854,0.0040904833,0.010317033],"study_design_scores_gemma":[0.00010370765,0.000028376693,0.0014173986,0.000072862495,0.000028914103,0.0000035012736,0.007287334,0.00035219343,0.00019242712,0.044209886,0.9460683,0.00023512334],"about_ca_topic_score_codex":0.0022285248,"about_ca_topic_score_gemma":0.006232953,"teacher_disagreement_score":0.9419778,"about_ca_system_score_codex":0.0008598411,"about_ca_system_score_gemma":0.00076032383,"threshold_uncertainty_score":0.9979599},"labels":[],"label_agreement":null},{"id":"W4412505613","doi":"10.70777/si.v2i3.15137","title":"Comparing Apples to Oranges: A Taxonomy for Navigating the Global Landscape of AI Regulation","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"","funders":"","keywords":"Taxonomy (biology); Geography; Biology; Ecology","score_opus":0.07006106803144764,"score_gpt":0.3828773494217818,"score_spread":0.3128162813903342,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4412505613","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.032623727,0.00045482855,0.86181825,0.07595792,0.0011288218,0.0036034454,0.00007627009,0.00010465783,0.024232082],"genre_scores_gemma":[0.9905366,0.00013612187,0.008006253,0.0008995444,0.0001421339,0.000083899846,0.000011155085,0.000008437936,0.00017589203],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.998279,0.00012983171,0.00047733853,0.0002639843,0.00043255888,0.00041732233],"domain_scores_gemma":[0.99871343,0.000442259,0.00010442602,0.0002661019,0.0003487314,0.00012504162],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0014792633,0.00015410676,0.00028401476,0.000028704366,0.00088491815,0.00015740625,0.0004936892,0.00012620608,0.000035250094],"category_scores_gemma":[0.00045887,0.00013346529,0.00013820245,0.00048777653,0.00029206145,0.00015295253,0.00013308822,0.00019367394,0.0000066777097],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00006708506,0.00007789544,0.026675032,0.000070337665,0.00009035674,3.6775137e-7,0.013688978,0.09936322,0.000116501156,0.8498672,0.0005003444,0.009482646],"study_design_scores_gemma":[0.0013526536,0.00066683965,0.022457657,0.002581702,0.00047497568,0.0000018846092,0.24776995,0.027401688,0.00997241,0.36825266,0.3172857,0.0017818843],"about_ca_topic_score_codex":0.0023082497,"about_ca_topic_score_gemma":0.0038549418,"teacher_disagreement_score":0.95791286,"about_ca_system_score_codex":0.00031136713,"about_ca_system_score_gemma":0.00036360644,"threshold_uncertainty_score":0.68061644},"labels":[],"label_agreement":null},{"id":"W4412505633","doi":"10.70777/si.v2i3.15063","title":"Darwin Gödel Machine: Open-Ended Evolution of Self-Improving Agents","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Evolutionary Algorithms and Applications","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Vector Institute; Canadian Institute for Advanced Research; University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada; Government of Canada; Canadian Institute for Advanced Research","keywords":"Darwin (ADL); Computer science; Software engineering","score_opus":0.017501471754553053,"score_gpt":0.28034707022915495,"score_spread":0.26284559847460187,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4412505633","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0006381384,0.00047698081,0.98963404,0.0017859992,0.0005485562,0.0008136421,0.000030892566,0.00015512809,0.005916605],"genre_scores_gemma":[0.6902934,0.00025921525,0.3083371,0.00018422543,0.000034181732,0.00007816588,0.000020549987,0.000014191617,0.0007790382],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99739313,0.00009314172,0.0008180709,0.0007510316,0.00047988506,0.0004647527],"domain_scores_gemma":[0.99817175,0.00009567905,0.00017557912,0.0012279561,0.00018662328,0.000142402],"candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0005214629,0.00028547324,0.00033762158,0.0001691227,0.00040893155,0.00014002167,0.002503368,0.000106936306,0.00007027271],"category_scores_gemma":[0.000033638415,0.00029277866,0.00013668682,0.00094070804,0.00010074625,0.00057665183,0.0016412474,0.00019723298,0.000058096117],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000017513898,0.00083278614,0.0009451335,0.00008521036,0.000114276954,0.0000051757706,0.0005003161,0.09974301,0.0025804946,0.87201864,0.0005345713,0.022622854],"study_design_scores_gemma":[0.00048889295,0.00016220604,0.0072418493,0.00012771724,0.000063230254,0.000014620993,0.0003885903,0.94661796,0.008902059,0.02680484,0.008665624,0.0005223814],"about_ca_topic_score_codex":0.0003715551,"about_ca_topic_score_gemma":0.000018142764,"teacher_disagreement_score":0.84687495,"about_ca_system_score_codex":0.000611991,"about_ca_system_score_gemma":0.00041349765,"threshold_uncertainty_score":0.99995244},"labels":[],"label_agreement":null},{"id":"W4412505644","doi":"10.70777/si.v2i3.15155","title":"Trends in Frontier AI Model Count: A Forecast to 2028","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Big Data and Business Intelligence","field":"Business, Management and Accounting","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Institute on Governance","funders":"","keywords":"Frontier; Econometrics; Computer science; Economics; Geography; Archaeology","score_opus":0.046812761081658584,"score_gpt":0.30773702969738564,"score_spread":0.26092426861572704,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4412505644","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.006181432,0.00036919268,0.9418161,0.013768002,0.0022292263,0.00082692486,0.000073696094,0.00020029744,0.034535117],"genre_scores_gemma":[0.9750209,0.0002393403,0.005648276,0.009851198,0.00041651735,0.00012852492,0.000220095,0.00007323486,0.008401906],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9964074,0.000016474582,0.000998596,0.0009790622,0.0006402787,0.00095819094],"domain_scores_gemma":[0.99858785,0.000041611176,0.00011719675,0.0009377371,0.0002541399,0.00006145594],"candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0005522512,0.00055858307,0.000582519,0.0011632955,0.00025296674,0.000417931,0.0011221708,0.00020294212,0.0011662196],"category_scores_gemma":[0.00008436209,0.0005497142,0.00019304425,0.0025242423,0.00013588778,0.0011693899,0.0008335178,0.00034560525,0.0007162672],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0003254541,0.0005105848,0.014662441,0.0002195729,0.00008253494,0.00003360992,0.00028842257,0.7146886,0.00029391077,0.14215563,0.025418222,0.10132099],"study_design_scores_gemma":[0.00043622823,0.000034416273,0.0038186105,0.00053166796,0.00012616947,0.0000041101475,0.00068338314,0.71126306,0.001063851,0.01589065,0.26498747,0.0011603695],"about_ca_topic_score_codex":0.0009054704,"about_ca_topic_score_gemma":0.0009434585,"teacher_disagreement_score":0.96883947,"about_ca_system_score_codex":0.00036025845,"about_ca_system_score_gemma":0.00009431527,"threshold_uncertainty_score":0.99974686},"labels":[],"label_agreement":null},{"id":"W4413130428","doi":"10.70777/si.v2i5.15503","title":"The Singapore Consensus on Global AI Safety Research Priorities","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Université de Montréal; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Pace; Autonomy; Trustworthiness; Software deployment; Public relations; Computer science; Political science; Business; Computer security; Law","score_opus":0.16269717274950973,"score_gpt":0.4808835490017747,"score_spread":0.318186376252265,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4413130428","genre_codex":"commentary","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.16854653,0.00948577,0.05252273,0.62610936,0.016988698,0.009548399,0.0002049767,0.00070977915,0.11588373],"genre_scores_gemma":[0.9860605,0.003938597,0.001251785,0.0027704046,0.00043729477,0.00007265815,0.00003461523,0.00003081055,0.0054033063],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","domain_scores_codex":[0.9954829,0.00037580205,0.0012138863,0.0006916339,0.0011246654,0.0011110909],"domain_scores_gemma":[0.9958703,0.0017035569,0.00008603079,0.0011288159,0.0008869183,0.00032440253],"candidate_categories":["metaepi_narrow","sts"],"consensus_categories":[],"category_scores_codex":[0.0022952678,0.00034341848,0.00042778844,0.00015277957,0.001753958,0.0001559232,0.00044584365,0.00025393948,0.00014976674],"category_scores_gemma":[0.0011401167,0.0002653014,0.00019807713,0.0009883613,0.0010122951,0.00005876857,0.00018836324,0.00085248944,0.00039201145],"study_design_candidate":"not_applicable","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0035670856,0.0010152945,0.020262014,0.0003235944,0.0003165886,0.00009573276,0.0030136106,0.017102245,0.00029856892,0.5835154,0.014741235,0.35574862],"study_design_scores_gemma":[0.0004670183,0.004327139,0.0139913,0.0032688822,0.00031850056,0.00019520629,0.074699685,0.011551974,0.06379456,0.12799594,0.6980184,0.0013714363],"about_ca_topic_score_codex":0.0009594879,"about_ca_topic_score_gemma":0.00036790932,"teacher_disagreement_score":0.817514,"about_ca_system_score_codex":0.001964097,"about_ca_system_score_gemma":0.0016755663,"threshold_uncertainty_score":0.9999799},"labels":[],"label_agreement":null},{"id":"W4413207531","doi":"10.70777/si.v2i5.15509","title":"In Which Areas of Technical AI Safety Could Geopolitical Rivals Cooperate?","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute; Institute on Governance","funders":"","keywords":"Harm; Geopolitics; Risk analysis (engineering); China; Business; Computer security; Political science; Computer science","score_opus":0.026117248207049245,"score_gpt":0.37977268411820253,"score_spread":0.3536554359111533,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4413207531","genre_codex":"other","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.046617128,0.001395341,0.10827548,0.34304985,0.0050031915,0.0046288883,0.00029590595,0.00044060012,0.4902936],"genre_scores_gemma":[0.9906809,0.0029173698,0.0012991141,0.0026145573,0.000140807,0.000020006704,0.000013730285,0.000024207156,0.0022893308],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.99569863,0.0004768972,0.0011987301,0.0005539489,0.0010318124,0.0010399552],"domain_scores_gemma":[0.99735326,0.00082740583,0.00010283596,0.0005430911,0.0008050187,0.00036840726],"candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.002899456,0.00031959996,0.0006906248,0.00019114827,0.0005723339,0.00013249616,0.0007181909,0.0006033666,0.00067514885],"category_scores_gemma":[0.0021271703,0.0003282781,0.00018831802,0.0012437372,0.0012616944,0.00037880745,0.00029717467,0.000805946,0.000047360634],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00010573173,0.0004327633,0.0031941533,0.000055236927,0.000054838492,0.00001819789,0.0050230618,0.00750428,0.0003587435,0.98223007,0.0004157167,0.0006072025],"study_design_scores_gemma":[0.0025688093,0.0011432488,0.028790595,0.0023787157,0.00042683596,0.00001792901,0.051477972,0.0038208861,0.021289105,0.5863668,0.2982924,0.0034267232],"about_ca_topic_score_codex":0.005155359,"about_ca_topic_score_gemma":0.0073839948,"teacher_disagreement_score":0.9440637,"about_ca_system_score_codex":0.00096541917,"about_ca_system_score_gemma":0.0016057128,"threshold_uncertainty_score":0.9999169},"labels":[],"label_agreement":null},{"id":"W4413417613","doi":"10.70777/si.v2i5.15569","title":"Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Space Science and Extraterrestrial Life","field":"Physics and Astronomy","cited_by":11,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"SAFER; Path (computing); Computer science; Artificial intelligence; Risk analysis (engineering); Computer security; Medicine","score_opus":0.029740865380178914,"score_gpt":0.3159395806666005,"score_spread":0.2861987152864216,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4413417613","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.2351901,0.00060714927,0.70848715,0.013686982,0.014931467,0.003896209,0.0013306092,0.00037077934,0.021499572],"genre_scores_gemma":[0.9927804,0.00013694217,0.001537198,0.0008080727,0.00043764355,0.00008408337,0.0002014807,0.000047256555,0.003966924],"study_design_codex":"simulation_or_modeling","study_design_gemma":"not_applicable","domain_scores_codex":[0.9952983,0.00015889885,0.0010737511,0.0012210651,0.0008840425,0.001363899],"domain_scores_gemma":[0.99787366,0.00008779901,0.00013920659,0.0011880241,0.00019540248,0.0005158925],"candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0006294249,0.00067492464,0.0006342855,0.0002823609,0.00073490915,0.00042393603,0.0011640227,0.0001296149,0.002451066],"category_scores_gemma":[0.00003068806,0.0006322396,0.00040601607,0.0010172037,0.0005106073,0.00031998998,0.00050977786,0.00046253446,0.000595393],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00075680396,0.003396174,0.22721244,0.00021817053,0.0013500438,0.00016338314,0.0067765596,0.39825615,0.011807308,0.26021528,0.019957004,0.06989064],"study_design_scores_gemma":[0.0061015934,0.002168883,0.055505406,0.002585726,0.0016631479,0.00005859929,0.04256151,0.07147599,0.2930809,0.06554618,0.44963816,0.009613909],"about_ca_topic_score_codex":0.0026388858,"about_ca_topic_score_gemma":0.00010500388,"teacher_disagreement_score":0.7575903,"about_ca_system_score_codex":0.00037808146,"about_ca_system_score_gemma":0.0005758604,"threshold_uncertainty_score":0.99961287},"labels":[],"label_agreement":null},{"id":"W4415233743","doi":"10.70777/si.v2i6.16169","title":"Responsible Agentic Reasoning and AI Agents: A Critical Survey","year":2025,"lang":"en","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Logic, Reasoning, and Knowledge","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Vector Institute","funders":"","keywords":"Operationalization; Intersection (aeronautics); Audit; Trustworthiness; Automated reasoning; Soundness; Class (philosophy)","score_opus":0.028270621759054662,"score_gpt":0.32028397629638267,"score_spread":0.292013354537328,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4415233743","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.003622492,0.0015985597,0.9837622,0.0029353942,0.0015231414,0.00040560725,0.000012845823,0.00022054407,0.0059192018],"genre_scores_gemma":[0.9534121,0.0010253558,0.04135245,0.0013111656,0.000063854866,0.000032072592,0.0000139528365,0.000027136668,0.0027618958],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.996265,0.0005263743,0.00070800906,0.0010814243,0.0005182147,0.00090097677],"domain_scores_gemma":[0.9972624,0.0009239914,0.000066794724,0.0011221466,0.00027482622,0.00034984227],"candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0017539929,0.00042231518,0.0004860887,0.00023691247,0.0005387939,0.00048236913,0.0011990181,0.00017666839,0.00012143401],"category_scores_gemma":[0.0012635208,0.00040681253,0.0001454772,0.0009092567,0.00033061174,0.00047919873,0.0011480028,0.00034573828,0.00016291137],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00010720317,0.00037115725,0.01693825,0.00013344083,0.0001323134,0.00011939472,0.0019204232,0.0046439515,0.00031035,0.96160793,0.0017348252,0.011980734],"study_design_scores_gemma":[0.0019532274,0.0013734229,0.18629098,0.0018807334,0.00033699963,0.00029622187,0.0017075596,0.5961468,0.02261992,0.13106999,0.05240444,0.003919701],"about_ca_topic_score_codex":0.00018890305,"about_ca_topic_score_gemma":0.00012583286,"teacher_disagreement_score":0.94978964,"about_ca_system_score_codex":0.00026870286,"about_ca_system_score_gemma":0.00038900352,"threshold_uncertainty_score":0.99983835},"labels":[],"label_agreement":null},{"id":"W4415454748","doi":"10.70777/si.v2i6.16253","title":"International Al Safety Report: First Key Update Capabilities and Risk Implications","year":2025,"lang":"","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Aluminum toxicity and tolerance in plants and animals","field":"Agricultural and Biological Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Toronto; Institute on Governance; Mila - Quebec Artificial Intelligence Institute","funders":"European Commission","keywords":"Key (lock); Field (mathematics); Information system; Information technology","score_opus":0.016276140037522657,"score_gpt":0.25653775122248174,"score_spread":0.2402616111849591,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4415454748","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.5805536,0.009690018,0.050765146,0.29122525,0.011079761,0.0041505317,0.0100347,0.0004383301,0.04206267],"genre_scores_gemma":[0.88622373,0.10711951,0.0013564277,0.001064412,0.00040173647,0.000045737073,0.0003196134,0.0000072777793,0.0034615546],"study_design_codex":"observational","study_design_gemma":"not_applicable","domain_scores_codex":[0.99544996,0.00016482273,0.0016683466,0.0013371329,0.00053202466,0.00084773666],"domain_scores_gemma":[0.99801266,0.0006263478,0.00039191934,0.00040948758,0.00024419682,0.00031540645],"candidate_categories":["metaepi_narrow","sts","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0009240177,0.0006265376,0.00067302136,0.000062927276,0.0018884484,0.0004365885,0.00089193817,0.00031289432,0.0025671988],"category_scores_gemma":[0.00020512899,0.0003744373,0.00032956764,0.00044629775,0.00090501196,0.00038826602,0.0006343348,0.0004950883,0.00009750757],"study_design_candidate":"not_applicable","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0013869429,0.0024633554,0.3780713,0.00030760976,0.0021234164,0.0002237323,0.0039223884,0.045331776,0.0041440534,0.21112716,0.007016246,0.343882],"study_design_scores_gemma":[0.00028583463,0.00029945705,0.12886874,0.0004272369,0.00022430923,0.00016340695,0.0035021072,0.004507986,0.0010432315,0.008148923,0.8515625,0.00096629316],"about_ca_topic_score_codex":0.0014178903,"about_ca_topic_score_gemma":0.001410075,"teacher_disagreement_score":0.8445462,"about_ca_system_score_codex":0.00027507465,"about_ca_system_score_gemma":0.00008710267,"threshold_uncertainty_score":0.9998708},"labels":[],"label_agreement":null},{"id":"W4415757988","doi":"10.70777/si.v2i5.16331","title":"A Grading Rubric for AI Safety Frameworks","year":2025,"lang":"","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Ethics and Social Impacts of AI","field":"Social Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Institute on Governance","funders":"","keywords":"Rubric; Grading (engineering); Warrant; Delphi method; Delphi","score_opus":0.0323729125467903,"score_gpt":0.3708621776750627,"score_spread":0.3384892651282724,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4415757988","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0005486151,0.0035905507,0.820213,0.1267953,0.010032379,0.0037956173,0.00025919187,0.00021902767,0.03454631],"genre_scores_gemma":[0.9229221,0.039415464,0.0132705495,0.012606499,0.0012000777,0.00014050993,0.000065502434,0.00014124265,0.010238058],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9904405,0.00073241984,0.0024581477,0.0017276393,0.0017743604,0.0028669508],"domain_scores_gemma":[0.9928394,0.0031505316,0.00045626622,0.0012487575,0.0013144056,0.0009906531],"candidate_categories":["metaepi_narrow","sts","scholarly_communication","research_integrity","insufficient_payload"],"consensus_categories":["research_integrity"],"category_scores_codex":[0.0053573293,0.0011169843,0.0015147127,0.00038556676,0.0052022566,0.0013446711,0.0021322374,0.0025644344,0.00093606726],"category_scores_gemma":[0.0028020202,0.0012540277,0.0012439457,0.0021486292,0.0020605177,0.00096989755,0.00068912416,0.0028222287,0.00015433852],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0003935279,0.0006536615,0.0007011672,0.0003861628,0.0007293353,0.000014974077,0.029583927,0.042694096,0.0001157383,0.89471763,0.002545314,0.027464462],"study_design_scores_gemma":[0.0011730428,0.0009360185,0.00051322015,0.0026053558,0.0010372951,0.0000039780325,0.049803298,0.021789059,0.0017355148,0.5255078,0.3921654,0.0027300469],"about_ca_topic_score_codex":0.0017263339,"about_ca_topic_score_gemma":0.0008915103,"teacher_disagreement_score":0.9223735,"about_ca_system_score_codex":0.002537464,"about_ca_system_score_gemma":0.0024032395,"threshold_uncertainty_score":0.99997723},"labels":[],"label_agreement":null},{"id":"W4417101019","doi":"10.70777/si.v2i4.16671","title":"International AI Safety Report 2025: Second Key Update: Technical Safeguards and Risk Management","year":2025,"lang":"","type":"article","venue":"SuperIntelligence - Robotics - Safety & Alignment","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"University of Toronto; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Key (lock); Risk management; SAFER; Risk assessment; Safeguard; Corporate governance; Risk governance","score_opus":0.009785592565935903,"score_gpt":0.2868525432628055,"score_spread":0.2770669506968696,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4417101019","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.00023911537,0.001197408,0.92787355,0.012652235,0.008797884,0.0016401637,0.00011823006,0.00030749198,0.047173906],"genre_scores_gemma":[0.48227653,0.04803244,0.4436655,0.0024191956,0.00083582243,0.0001538408,0.00024495035,0.00019387752,0.022177834],"study_design_codex":"simulation_or_modeling","study_design_gemma":"not_applicable","domain_scores_codex":[0.98846745,0.0007138882,0.003545367,0.0034144032,0.002157464,0.001701432],"domain_scores_gemma":[0.99441737,0.0004904865,0.0008444645,0.0031517467,0.00050279335,0.0005931635],"candidate_categories":["metaepi_narrow","sts","insufficient_payload"],"consensus_categories":["metaepi_narrow"],"category_scores_codex":[0.003977937,0.001388208,0.0013223029,0.0007069146,0.0014522431,0.0010009899,0.004015722,0.0006407808,0.0026646],"category_scores_gemma":[0.0004464759,0.0015355045,0.00054378243,0.0015558124,0.0010082058,0.0011803131,0.007950736,0.0022427207,0.00026459675],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00044311045,0.00050834933,0.0059988685,0.00027784467,0.0014140095,0.0012583632,0.00077796064,0.47015017,0.000047603015,0.33204922,0.001494022,0.18558048],"study_design_scores_gemma":[0.0016542434,0.00032594183,0.01208614,0.0010366943,0.0008193096,0.00065271434,0.0010119019,0.43308887,0.00083467015,0.014804647,0.531373,0.0023118525],"about_ca_topic_score_codex":0.00016789557,"about_ca_topic_score_gemma":0.00009254226,"teacher_disagreement_score":0.52987903,"about_ca_system_score_codex":0.0019523709,"about_ca_system_score_gemma":0.0005244201,"threshold_uncertainty_score":0.9998868},"labels":[],"label_agreement":null}]}