{"id":"W3167967639","doi":"","title":"LEGO: Latent Execution-Guided Reasoning for Multi-Hop Question Answering on Knowledge Graphs","year":2021,"lang":"en","type":"article","venue":"International Conference on Machine Learning","topic":"Topic Modeling","field":"Computer Science","cited_by":31,"is_retracted":false,"has_abstract":false,"ca_institutions":"Google (Canada); University of Alberta","funders":"","keywords":"Computer science; Question answering; Knowledge graph; Hop (telecommunications); Natural language processing; Artificial intelligence; Computer network","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001781271,0.001967999,0.001332373,0.00255164,0.001124381,0.003298455,0.003312114,0.002236361,0.01465998],"category_scores_gemma":[0.009394085,0.001013056,0.002844297,0.001510385,0.001096754,0.005954714,0.005066518,0.003060741,0.004039806],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001402124,"about_ca_system_score_gemma":0.002015936,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01056524,"about_ca_topic_score_gemma":0.02312426,"domain_scores_codex":[0.9982174,0.0005863425,0.0001101601,0.0004786583,0.0004403316,0.0001670416],"domain_scores_gemma":[0.9959777,0.00263644,0.0001557459,0.0008173229,0.0002673301,0.0001454211],"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":[0.001612589,0.0009372862,0.004189725,0.001749239,0.0006011173,0.0008021699,0.001474806,0.1243337,0.01571962,0.09183411,0.1057148,0.6510309],"study_design_scores_gemma":[0.0001459494,0.00007671155,0.0003496112,0.00008384437,0.0001020554,0.00007383711,0.0001869171,0.8627472,0.004283352,0.1198888,0.01202237,0.00003941049],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009475557,0.0004785687,0.918133,0.0006091434,0.0001310165,0.00031829,0.004808764,0.06302729,0.003018327],"genre_scores_gemma":[0.2033065,0.000361783,0.7703645,0.0005279301,0.0001033794,0.0004032013,0.01658212,0.003286519,0.00506415],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01465998,"threshold_uncertainty_score":0.04904246,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08323708750218646,"score_gpt":0.3583986716908126,"score_spread":0.2751615841886262,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}