{"id":"W4415283783","doi":"10.1177/20539517251386055","title":"Cosine capital: Large language models and the embedding of all things","year":2025,"lang":"en","type":"article","venue":"Big Data & Society","topic":"Complex Systems and Time Series Analysis","field":"Economics, Econometrics and Finance","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"Koneen Säätiö","keywords":"Commodification; Abstraction; Embedding; sort; Language model; Process (computing); Modeling language; Natural language","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":["sts"],"consensus_categories":[],"category_scores_codex":[0.00314068,0.0008843212,0.0009132996,0.001657968,0.0013621,0.006305972,0.001398289,0.001542896,0.006726633],"category_scores_gemma":[0.02684965,0.0005807293,0.001042842,0.001754992,0.005948703,0.01335071,0.004477615,0.003411756,0.0007871411],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001956337,"about_ca_system_score_gemma":0.00117989,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004320842,"about_ca_topic_score_gemma":0.003101427,"domain_scores_codex":[0.9975972,0.001315125,0.00009528327,0.0004572191,0.0003660735,0.0001692477],"domain_scores_gemma":[0.9865317,0.009129626,0.001229492,0.001958108,0.0006227345,0.0005283676],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0000265183,0.00001038242,0.0007073336,0.00003157485,0.00002562967,0.00006395399,0.0003282265,0.01026917,0.000118869,0.9783176,0.001392956,0.008707757],"study_design_scores_gemma":[0.000005087203,0.000007947905,0.0001617714,0.00001790026,0.000005880398,0.00003617411,0.00004483196,0.0507657,0.00007562202,0.946107,0.002760546,0.00001157114],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05753589,0.002651656,0.8988018,0.01318287,0.0003424069,0.00006145659,0.00083672,0.0004121103,0.02617515],"genre_scores_gemma":[0.9010984,0.00214631,0.0827125,0.001377232,0.0006832924,0.0001925857,0.0007342948,0.0002421509,0.01081324],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9986379,"threshold_uncertainty_score":0.02250278,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06613855722414949,"score_gpt":0.2689227084585134,"score_spread":0.2027841512343639,"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."}}