{"paper":{"title":"Where Should Learning Live? Regret Analysis for Text, Parameter, and Joint Adaptation","arxiv_id":"2611.09701","primary_cat":"stat.ML","submitted_at":"2026-11-06T16:46:11+00:00","version_compiled":"v1"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v1","sentences":[{"i":0,"p":1.0,"id":"s0","own":false,"role":"background","text":"Language-model systems can select prompts and parameter adapters repeatedly during deployment, using feedback to guide later choices.","hedge":{"p":0.6,"label":"asserted"},"own_p":0.09,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":1.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":1.0,"id":"s1","own":false,"role":"background","text":"Earlier choices may change the context or environment inherited by later configurations.","hedge":{"p":0.94,"label":"hedged_possibility"},"own_p":0.11,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":1.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.91,"id":"s2","own":true,"role":"contribution","text":"We study selection among preconstructed candidates and compare cumulative loss with the best fixed prompt--parameter pair used from the beginning.","hedge":{"p":0.94,"label":"asserted"},"own_p":0.84,"main_claim_p":0.05,"quantitative":true,"quantitative_p":0.64,"role_probabilities":{"gap":0.0,"method":0.09,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.91,"interpretation":0.0}},{"i":3,"p":0.43,"id":"s3","own":true,"role":"contribution","text":"This excess loss separates into three terms: selection regret within the available family, the gap from restricting that family, and the benefit or harm of the actual deployment history.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.84,"main_claim_p":0.16,"quantitative":false,"quantitative_p":0.45,"role_probabilities":{"gap":0.0,"method":0.32,"result":0.19,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.43,"interpretation":0.06}},{"i":4,"p":0.97,"id":"s4","own":true,"role":"method","text":"Selection is evaluated using the history each candidate would have produced if used throughout, which separates candidate quality from the effects of switching.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.8,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.1,"role_probabilities":{"gap":0.0,"method":0.97,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.03,"interpretation":0.0}},{"i":5,"p":0.85,"id":"s5","own":true,"role":"result","text":"When loss functions are fixed independently of selector randomness, past choices affect loss for only a bounded number of rounds, and every candidate's own-history loss is revealed, limiting switches yields sublinear expected regret within the selected family.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.74,"main_claim_p":0.23,"quantitative":true,"quantitative_p":0.82,"role_probabilities":{"gap":0.0,"method":0.04,"result":0.85,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.09,"interpretation":0.02}},{"i":6,"p":0.88,"id":"s6","own":true,"role":"contribution","text":"We also quantify the costs and additional assumptions of independent probing and deployment-only feedback.","hedge":{"p":0.97,"label":"asserted"},"own_p":0.85,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.05,"role_probabilities":{"gap":0.0,"method":0.1,"result":0.0,"background":0.0,"limitation":0.02,"future_work":0.0,"contribution":0.88,"interpretation":0.0}},{"i":7,"p":0.98,"id":"s7","own":true,"role":"result","text":"Controlled experiments show that fewer switches can either lower or raise deployed loss, depending on whether inherited histories are harmful or helpful.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.94,"main_claim_p":0.23,"quantitative":true,"quantitative_p":0.79,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.98,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.02}},{"i":8,"p":0.99,"id":"s8","own":true,"role":"result","text":"With Qwen3.5-9B on Spider, text-only selection has lower within-family regret than joint selection, yet produces more failures because its candidate family excludes stronger configurations.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.95,"main_claim_p":0.26,"quantitative":true,"quantitative_p":0.96,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.99,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.01}}]},"compiled":{"main_claim":{"p":0.57,"text":"Controlled experiments show that fewer switches can either lower or raise deployed loss, depending on whether inherited histories are harmful or helpful.","location":"abstract s7","hedge_strength":{"p":0.98,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":6822,"cost_usd_reported":0.0},"compiled_at":"2026-11-09","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.04,"states_limitation":0.87,"human_participants":0.04,"controlled_comparison":0.89,"randomized_assignment":0.04,"states_quantitative_result":0.23,"compares_with_prior_results":0.88,"mentions_code_or_data_availability":0.03},"evidential_basis":{"p":0.77,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.22,"analysis_of_existing_data":0.0,"observation_or_experiment":0.77,"simulation_or_computation":0.01}},"contribution_type":{"p":0.92,"value":"theoretical result","probabilities":{"review":0.0,"method_or_tool":0.01,"empirical_study":0.07,"simulation_study":0.0,"theoretical_result":0.92,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":0.85,"text":"We study selection among preconstructed candidates and compare cumulative loss with the best fixed prompt--parameter pair used from the beginning.","location":"abstract s2"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v1"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":9,"sentences_in_abstract":9},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}