{"id":"W3015648683","doi":"10.18653/v1/2021.eacl-main.246","title":"ProFormer: Towards On-Device LSH Projection Based Transformers","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Memory footprint; Computer science; Embedding; Computation; Projection (relational algebra); Precomputation; Mobile device; Dimension (graph theory); Transformer; Word embedding; Theoretical computer science; Artificial intelligence; Speech recognition; Algorithm; Mathematics; Programming language; Engineering","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":[0.0007571241,0.0008901792,0.0006684987,0.0005435834,0.0002763605,0.001414618,0.001942432,0.0006714007,0.01673448],"category_scores_gemma":[0.003227972,0.0005640956,0.000568077,0.0007525384,0.0006234893,0.004559783,0.002248432,0.001629189,0.007565589],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005710759,"about_ca_system_score_gemma":0.000957071,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002186076,"about_ca_topic_score_gemma":0.004195065,"domain_scores_codex":[0.9994655,0.0001070512,0.00003534433,0.0001404079,0.0001958766,0.00005580423],"domain_scores_gemma":[0.9992317,0.0002405482,0.00003422622,0.0002668159,0.0001809011,0.0000458025],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007828398,0.0002773351,0.0007862711,0.0003007952,0.0000853825,0.0002260816,0.0002024599,0.02821043,0.05490889,0.0291239,0.03087306,0.8542227],"study_design_scores_gemma":[0.00009166636,0.0002365102,0.0003256702,0.00003047768,0.00003327708,0.0002710616,0.00009876378,0.8729817,0.05850416,0.04751389,0.01987329,0.00003951623],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01262576,0.000322639,0.9604701,0.0002438536,0.0001617341,0.0001293486,0.0005572165,0.02151797,0.003971305],"genre_scores_gemma":[0.4140085,0.0007100683,0.5602961,0.0007396947,0.0001721372,0.000339273,0.003233599,0.001912572,0.018588],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01673448,"threshold_uncertainty_score":0.05598241,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04861438386620129,"score_gpt":0.2893038014633784,"score_spread":0.2406894175971771,"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."}}