{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01157424,0.003788173,0.002043193,0.002196338,0.001640039,0.002871971,0.00415376,0.00385501,0.005146584],"category_scores_gemma":[0.04095657,0.0008651848,0.001335628,0.001518502,0.001479174,0.006428629,0.004840584,0.005017275,0.00270883],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002717453,"about_ca_system_score_gemma":0.002213332,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01676529,"about_ca_topic_score_gemma":0.02425356,"domain_scores_codex":[0.9900283,0.005542515,0.0005555543,0.002091121,0.001310782,0.0004717504],"domain_scores_gemma":[0.9837307,0.01074112,0.0005095507,0.002324828,0.001929795,0.0007640757],"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":[0.002288902,0.001835894,0.007783035,0.001578875,0.000886848,0.0004906967,0.00050211,0.2942561,0.01434691,0.004694388,0.03211839,0.6392179],"study_design_scores_gemma":[0.0001812907,0.001202149,0.001944821,0.0001461333,0.0001008816,0.0002578303,0.0003324395,0.9721819,0.01165523,0.007584823,0.004299928,0.0001125332],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3221135,0.02331111,0.5671146,0.002740711,0.002140026,0.001604185,0.006224188,0.05650939,0.01824234],"genre_scores_gemma":[0.7152879,0.001178932,0.261374,0.001590489,0.0002558958,0.0007735642,0.01312698,0.001683752,0.004728541],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01676529,"threshold_uncertainty_score":0.06121117,"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."}}