{"id":"W3206282310","doi":"10.18653/v1/2023.eacl-main.55","title":"What Makes Sentences Semantically Related? A Textual Relatedness Dataset and Empirical Study","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; National Research Council Canada; Institute for Work & Health","funders":"","keywords":"Computer science; Natural language processing; Semantic similarity; Annotation; Artificial intelligence; Automatic summarization; Intuition; Sentence; Similarity (geometry); Information retrieval; Psychology","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.001042375,0.0003854042,0.0005098347,0.0002227325,0.0001499763,0.001628922,0.001819009,0.0003788677,0.00002495531],"category_scores_gemma":[0.0001649184,0.0003145676,0.00006300192,0.0003874966,0.000108486,0.0008012792,0.007740064,0.001035503,0.0001871693],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003704934,"about_ca_system_score_gemma":0.0001943169,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000320728,"about_ca_topic_score_gemma":0.0002496511,"domain_scores_codex":[0.9961464,0.0002602846,0.0007206392,0.001754375,0.0006815334,0.0004368043],"domain_scores_gemma":[0.9974062,0.0003100983,0.000156619,0.001818582,0.0001117577,0.0001967352],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001448311,0.006235973,0.3100193,0.002180686,0.004675698,0.01570778,0.1315431,0.02654552,0.00007598337,0.08951391,0.2261409,0.1872163],"study_design_scores_gemma":[0.001006167,0.0003029379,0.01868647,0.000789443,0.0001736364,0.00019743,0.01400166,0.9231263,0.00001060729,0.03837518,0.001797568,0.001532607],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6470078,0.0009877912,0.3071132,0.03034222,0.008555111,0.002508042,0.0001968983,0.002564932,0.0007240027],"genre_scores_gemma":[0.9640711,0.0008550257,0.02847602,0.001019018,0.0002022195,0.000116795,0.0005404311,0.00006965292,0.004649776],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8965808,"threshold_uncertainty_score":0.9999306,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09582439625974269,"score_gpt":0.3546847947762765,"score_spread":0.2588603985165338,"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."}}