{"id":"W4386566893","doi":"10.18653/v1/2023.eacl-main.49","title":"Combining Parameter-efficient Modules for Task-level Generalisation","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Mila - Quebec Artificial Intelligence Institute; McGill University","funders":"Institut de Valorisation des Données; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Reinforcement learning; Modular design; Benchmark (surveying); Machine learning; Artificial intelligence; Task (project management); Latent variable; Language model","routes":{"ca_aff":true,"ca_fund":true,"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":[0.002675555,0.00196702,0.001223913,0.0007965686,0.0004659749,0.001095264,0.003047839,0.002012291,0.004117711],"category_scores_gemma":[0.01015945,0.0009449134,0.001432963,0.0007701897,0.001599869,0.004383564,0.003790195,0.003859409,0.001938021],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001132043,"about_ca_system_score_gemma":0.001065622,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002839242,"about_ca_topic_score_gemma":0.00544526,"domain_scores_codex":[0.9989893,0.0003344419,0.00004955226,0.0003838346,0.0001293932,0.0001134647],"domain_scores_gemma":[0.9968118,0.001358569,0.0001854314,0.001157347,0.000316768,0.0001701345],"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":[0.0003569947,0.0003890773,0.004093911,0.0002269066,0.0002671791,0.0001612624,0.0003955986,0.6010508,0.02091285,0.01478603,0.003712088,0.3536474],"study_design_scores_gemma":[0.00002827631,0.00009380189,0.0002657746,0.00001149644,0.00003214648,0.00003541034,0.00001907332,0.9801326,0.002816531,0.0158066,0.0007450824,0.00001319513],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04460303,0.0002732131,0.9485478,0.0003152212,0.00003635154,0.0001298269,0.0001087571,0.004189188,0.00179653],"genre_scores_gemma":[0.7108456,0.0002457175,0.282375,0.0005494267,0.00008853106,0.0005564099,0.0006449342,0.0005780795,0.00411619],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004117711,"threshold_uncertainty_score":0.01414984,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1278035831734825,"score_gpt":0.2946614849743965,"score_spread":0.1668579018009139,"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."}}