{"id":"W7011290330","doi":"","title":"Low-Bit Power-of-Two Quantization for Large Language Models","year":2025,"lang":"en","type":"dissertation","venue":"eScholarship@McGill (McGill)","topic":"History of Computing Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Quantization (signal processing); Language model; Natural language; Language identification","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"],"consensus_categories":[],"category_scores_codex":[0.0009687911,0.0007039396,0.0009190811,0.001049299,0.0008864711,0.0001145535,0.003276938,0.0008365298,0.00001555206],"category_scores_gemma":[0.001048807,0.0008134918,0.0004927856,0.001122011,0.00003959349,0.001100032,0.0005077891,0.0009149582,0.00004075078],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005211888,"about_ca_system_score_gemma":0.0001798929,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006923391,"about_ca_topic_score_gemma":0.000443603,"domain_scores_codex":[0.9958027,0.0001966278,0.001051948,0.001364827,0.0007757644,0.0008081202],"domain_scores_gemma":[0.995869,0.0004303693,0.0009580919,0.001878153,0.0007361199,0.0001282785],"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":[0.00004468087,0.0002005833,0.000001612656,0.0005534079,0.00009935387,0.00001813772,0.00007233116,0.00042612,0.008246494,0.920821,0.00004058806,0.06947567],"study_design_scores_gemma":[0.003587504,0.0006196407,0.00006458244,0.002671982,0.0002690183,0.00001891227,0.0006468305,0.04674904,0.3647877,0.5410001,0.03665042,0.002934316],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8562165,0.002384821,0.03218239,0.0001341082,0.009431195,0.003892489,0.002164506,0.005953856,0.08764006],"genre_scores_gemma":[0.9407769,0.00004645664,0.05314038,0.0001255208,0.00002190037,0.0001443496,0.0004942125,0.0001070488,0.00514327],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3798209,"threshold_uncertainty_score":0.9994316,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01764006618815532,"score_gpt":0.2677293591335536,"score_spread":0.2500892929453983,"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."}}