{"id":"W4389519544","doi":"10.18653/v1/2023.newsum-1.12","title":"Analyzing Multi-Sentence Aggregation in Abstractive Summarization via the Shapley Value","year":2023,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Canadian Institute for Advanced Research; Nvidia","keywords":"Automatic summarization; Shapley value; Computer science; Sentence; Value (mathematics); Natural language processing; Artificial intelligence; Mathematics; Game theory; Machine learning; Mathematical economics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004496317,0.00009744793,0.0001049008,0.0002867911,0.0001128967,0.00009252543,0.0005966779,0.00004193478,0.000008366227],"category_scores_gemma":[0.0001183826,0.00007202004,0.00004979118,0.002257132,0.00003346022,0.0009140159,0.0001945464,0.0001282688,0.00008570684],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007831848,"about_ca_system_score_gemma":0.00002014245,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003291526,"about_ca_topic_score_gemma":0.0004786024,"domain_scores_codex":[0.9989725,0.00007332826,0.0002332815,0.0003230454,0.000198328,0.0001995192],"domain_scores_gemma":[0.9991934,0.000172785,0.0001162747,0.0004082707,0.00008242139,0.0000268191],"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":[0.00001118787,0.0002328348,0.08283867,0.00002145435,0.00008612593,0.0000551835,0.003862323,0.07403051,0.04565767,0.1625847,0.0005677929,0.6300516],"study_design_scores_gemma":[0.00008208282,0.000008137018,0.04903337,0.00001448684,0.000003589085,0.00000118273,0.0000636818,0.9284431,0.01250766,0.009669714,0.00007267464,0.0001003709],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01059969,0.00001910551,0.9874541,0.0008017593,0.00003886528,0.0001682493,3.536776e-7,0.0004181825,0.0004996569],"genre_scores_gemma":[0.9098688,0.00006474785,0.08950881,0.0001150075,0.00001552791,0.00003044883,0.00000874015,0.000007045155,0.0003808134],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8992692,"threshold_uncertainty_score":0.2936892,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02542860140765619,"score_gpt":0.3057754320863267,"score_spread":0.2803468306786705,"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."}}