{"paper":{"title":"Same-Number Citation Swaps: Stress-Testing Jev as a Financial Evidence Judge","arxiv_id":"2610.08675","primary_cat":"cs.CL","submitted_at":"2026-10-06T16:57:47+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":0.97,"id":"s0","own":false,"role":"background","text":"Financial reports repeat values across periods, metrics and accounting lines, allowing an LLM-generated calculation to be numerically correct while citing the wrong financial role.","hedge":{"p":0.91,"label":"asserted"},"own_p":0.14,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.09,"role_probabilities":{"gap":0.03,"method":0.0,"result":0.0,"background":0.97,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.83,"id":"s1","own":true,"role":"contribution","text":"We evaluate what probabilistic evidence verification adds beyond number matching using Jev as a source-support verifier for GPT-4.1-mini calculation traces.","hedge":{"p":0.75,"label":"asserted"},"own_p":0.91,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.21,"role_probabilities":{"gap":0.0,"method":0.17,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.83,"interpretation":0.0}},{"i":2,"p":0.97,"id":"s2","own":true,"role":"result","text":"A signed-number-at-pointer baseline explains most recovery over exact quotation checks.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.76,"main_claim_p":0.17,"quantitative":true,"quantitative_p":0.72,"role_probabilities":{"gap":0.0,"method":0.01,"result":0.97,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.02}},{"i":3,"p":0.99,"id":"s3","own":true,"role":"method","text":"To isolate the remaining role-recognition problem, we hold operands and arithmetic fixed, move citations between same-number cells, and retain controls that express equivalent facts.","hedge":{"p":0.78,"label":"asserted"},"own_p":0.93,"main_claim_p":0.05,"quantitative":false,"quantitative_p":0.15,"role_probabilities":{"gap":0.0,"method":0.99,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":4,"p":0.87,"id":"s4","own":true,"role":"result","text":"These contrasts reveal both wrong-role citations that pass and valid alternative citations that are withheld.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.91,"main_claim_p":0.23,"quantitative":false,"quantitative_p":0.27,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.87,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.13}},{"i":5,"p":1.0,"id":"s5","own":true,"role":"result","text":"Explicit column labels improve selected wrong-role decisions while also lowering support for some equivalent evidence.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.9,"main_claim_p":0.14,"quantitative":true,"quantitative_p":0.51,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":6,"p":0.75,"id":"s6","own":true,"role":"method","text":"A constructed follow-up on 36 new source pages, labeled by a non-author reviewer, extends this evaluation and exposes the same tradeoff between detecting role errors and retaining valid citations.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.87,"main_claim_p":0.12,"quantitative":true,"quantitative_p":0.96,"role_probabilities":{"gap":0.0,"method":0.75,"result":0.09,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.16,"interpretation":0.0}},{"i":7,"p":1.0,"id":"s7","own":true,"role":"contribution","text":"The contribution is a controlled evaluation that identifies what a probabilistic financial verifier distinguishes when numerical matching is held fixed.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.89,"main_claim_p":0.45,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":1.0,"interpretation":0.0}},{"i":8,"p":0.58,"id":"s8","own":false,"role":"interpretation","text":"For LLM-based financial assistants, it makes numerical correctness, cited-role support and acceptance outcomes separately assessable.","hedge":{"p":0.96,"label":"asserted"},"own_p":0.49,"main_claim_p":0.17,"quantitative":false,"quantitative_p":0.08,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.01,"contribution":0.41,"interpretation":0.58}}]},"compiled":{"main_claim":{"p":0.51,"text":"The contribution is a controlled evaluation that identifies what a probabilistic financial verifier distinguishes when numerical matching is held fixed.","location":"abstract s7","hedge_strength":{"p":0.99,"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":6787,"cost_usd_reported":0.0},"compiled_at":"2026-10-07","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.06,"states_limitation":0.79,"human_participants":0.49,"controlled_comparison":0.93,"randomized_assignment":0.03,"states_quantitative_result":0.52,"compares_with_prior_results":0.91,"mentions_code_or_data_availability":0.04},"evidential_basis":{"p":0.9,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.0,"analysis_of_existing_data":0.0,"observation_or_experiment":0.9,"simulation_or_computation":0.1}},"contribution_type":{"p":1.0,"value":"empirical study","probabilities":{"review":0.0,"method_or_tool":0.0,"empirical_study":1.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":0.59,"text":"The contribution is a controlled evaluation that identifies what a probabilistic financial verifier distinguishes when numerical matching is held fixed.","location":"abstract s7"}},"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"}}}