{"id":"W4412889920","doi":"10.18653/v1/2025.acl-long.1245","title":"On Many-Shot In-Context Learning for Long-Context Evaluation","year":2025,"lang":"en","type":"article","venue":"","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institute for Catastrophic Loss Reduction; National Science Foundation","keywords":"Shot (pellet); Computer science; Context (archaeology); One shot; Artificial intelligence; Human–computer interaction; Engineering; History","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00119652,0.000121586,0.0001470096,0.0003028183,0.0001553963,0.0001787474,0.0003740364,0.00006245522,0.0001579393],"category_scores_gemma":[0.000498655,0.0001166183,0.00007016327,0.000386609,0.00001833153,0.0003096829,0.00007085173,0.000195855,0.0001233054],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001217787,"about_ca_system_score_gemma":0.0001206288,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003084109,"about_ca_topic_score_gemma":0.0002126614,"domain_scores_codex":[0.9986548,0.000158014,0.000264826,0.0003961254,0.0002749819,0.0002512429],"domain_scores_gemma":[0.9990047,0.0004792986,0.00007335884,0.0002618533,0.00013696,0.00004388514],"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":[0.00002722207,0.00003888286,0.001234198,0.00001040154,0.000009541081,0.000001362392,0.0007730533,0.007139136,0.0000686926,0.4257818,0.0008152867,0.5641004],"study_design_scores_gemma":[0.001926578,0.0001324227,0.01660715,0.00007691118,0.000006868032,0.000001066158,0.0007756974,0.9529561,0.0005252202,0.01089007,0.01592155,0.0001803248],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02414732,0.0001247301,0.9080606,0.001854001,0.0004221811,0.0005308323,1.28485e-7,0.0001354282,0.06472476],"genre_scores_gemma":[0.9805524,0.000003656503,0.00330456,0.002654966,0.00001939871,0.0001018669,0.000005189155,0.000006918956,0.01335101],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9564051,"threshold_uncertainty_score":0.4755557,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04640130837711125,"score_gpt":0.3282955049327902,"score_spread":0.281894196555679,"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."}}