{"id":"W4382568066","doi":"10.1109/tkde.2023.3290792","title":"Individual and Structural Graph Information Bottlenecks for Out-of-Distribution Generalization","year":2023,"lang":"en","type":"article","venue":"IEEE Transactions on Knowledge and Data Engineering","topic":"Advanced Graph Neural Networks","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute","funders":"National Natural Science Foundation of China","keywords":"Spurious relationship; Computer science; Graph; Leverage (statistics); Artificial intelligence; Bottleneck; Theoretical computer science; Algorithm; Machine learning","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":[0.007690074,0.001946186,0.003387941,0.002842629,0.001894024,0.002981461,0.004857819,0.003227426,0.005885115],"category_scores_gemma":[0.04607046,0.001243467,0.001901632,0.00254475,0.002767429,0.009271656,0.004671907,0.005875935,0.001415141],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00425962,"about_ca_system_score_gemma":0.00291763,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009249385,"about_ca_topic_score_gemma":0.009629281,"domain_scores_codex":[0.996067,0.001177673,0.0001802519,0.00142105,0.0006921199,0.0004619799],"domain_scores_gemma":[0.9680889,0.02261622,0.001841594,0.004548926,0.001728352,0.001176081],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0008491551,0.0006700482,0.007190689,0.00062723,0.0002741691,0.0005648226,0.0006341696,0.4915234,0.004340774,0.1804872,0.03314465,0.2796936],"study_design_scores_gemma":[0.00002529639,0.00004357145,0.0006798275,0.0000237417,0.00002281264,0.00009210979,0.00005158693,0.8766367,0.001152891,0.1196036,0.0016492,0.0000187312],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04952684,0.001516977,0.9371938,0.002977627,0.0001934932,0.0001445032,0.0004744592,0.002841866,0.005130517],"genre_scores_gemma":[0.8212548,0.001583359,0.1628481,0.00227711,0.0005442051,0.0003791443,0.001935279,0.001266836,0.007911167],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009249385,"threshold_uncertainty_score":0.0406695,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02836388878460003,"score_gpt":0.2753856377479403,"score_spread":0.2470217489633403,"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."}}