{"id":"W7106831009","doi":"10.48448/fmry-df15","title":"NileChat: Towards Linguistically Diverse and Culturally Aware LLMs for Local Communities","year":2025,"lang":"","type":"other","venue":"Open MIND","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; University of British Columbia","funders":"","keywords":"Cultural diversity; Linguistic diversity; Inclusion (mineral); Work (physics); Cultural heritage; Translation (biology); Cultural values","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":[0.002459604,0.001411391,0.0007286387,0.001362529,0.001091249,0.002227899,0.002920442,0.001733756,0.005476617],"category_scores_gemma":[0.009128527,0.0005429504,0.001453672,0.0008419991,0.0006912292,0.00399202,0.004228121,0.001863506,0.004430447],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001610708,"about_ca_system_score_gemma":0.0019616,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02132074,"about_ca_topic_score_gemma":0.0399373,"domain_scores_codex":[0.9986013,0.0006911704,0.00007473343,0.0003243531,0.0002008725,0.00010763],"domain_scores_gemma":[0.9978472,0.0007451822,0.0001184806,0.0004958876,0.0005226056,0.0002707302],"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":[0.00125808,0.001135847,0.01774746,0.001189952,0.0007610001,0.0009682503,0.002343621,0.4784056,0.0208482,0.01432141,0.06374546,0.3972751],"study_design_scores_gemma":[0.00009533923,0.0001445033,0.0008917599,0.00008733873,0.00006192441,0.0001129385,0.0008248339,0.9600833,0.005010056,0.01403219,0.01860602,0.00004975429],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1387254,0.001131244,0.774981,0.002585441,0.0004282456,0.0008906702,0.008665017,0.05857856,0.0140143],"genre_scores_gemma":[0.5007328,0.0003559547,0.4578778,0.001063355,0.00009314297,0.0009524433,0.02641545,0.002839972,0.009669094],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02132074,"threshold_uncertainty_score":0.04239327,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06112835526601764,"score_gpt":0.3495471824085541,"score_spread":0.2884188271425365,"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."}}