{"id":"W7116175540","doi":"10.3103/s1060992x25601733","title":"Memory Stream: Enhancing Information Flow in Recurrent Memory Transformers for Efficient Long-Context Training","year":2025,"lang":"en","type":"article","venue":"Optical Memory and Neural Networks","topic":"Generative Adversarial Networks and Image Synthesis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Transformer; High memory; Memory model; Computational complexity theory; Flat memory model; Content-addressable memory; Matching (statistics); Architecture","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":[0.0005789238,0.0006785972,0.0005052798,0.0003036727,0.0002064311,0.000467359,0.00128235,0.0006279681,0.003751942],"category_scores_gemma":[0.001961197,0.0003052651,0.0004815884,0.0003042966,0.0004592087,0.001340728,0.001032984,0.001160281,0.000646393],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000417037,"about_ca_system_score_gemma":0.0006550864,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001879525,"about_ca_topic_score_gemma":0.003447518,"domain_scores_codex":[0.9998477,0.00003810282,0.000008898192,0.00004442605,0.00003385963,0.00002704204],"domain_scores_gemma":[0.9995536,0.0002291274,0.00004285284,0.00006984119,0.00007250944,0.00003188549],"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":[0.0003617443,0.0002059286,0.001129212,0.0001640354,0.00006940431,0.0001446604,0.0001253082,0.5660442,0.04279637,0.01461674,0.003255622,0.3710868],"study_design_scores_gemma":[0.000007945408,0.00004123052,0.00005085646,0.000004222275,0.000006432266,0.00001576306,0.000004914314,0.991151,0.006038289,0.002378742,0.000297342,0.000003228675],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05489569,0.000362923,0.9400774,0.0001749101,0.00006142113,0.00005549884,0.00008606829,0.002081931,0.002204185],"genre_scores_gemma":[0.8431284,0.0001905452,0.1526536,0.0001808341,0.00003073982,0.00009676327,0.0002230696,0.0001892523,0.003306588],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003751942,"threshold_uncertainty_score":0.01255149,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01365455953999392,"score_gpt":0.2390468300254826,"score_spread":0.2253922704854887,"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."}}