1[2 {"id":"C1","text":"Exactly one of the five references, R3, is unverifiable; the other four pass all three checks.","results_keys":["unverifiable_references","n_verified","reference_checks"],"status":"supported","caveat":"Judgement is restricted to the provided extracts."},3 {"id":"C2","text":"R3 fails all three checks: the entry's title differs from the document title, its year (2024) matches neither the v1 nor the v2 submission year (2026), and the extract reports a gradual rise from about 0.5% in the event year to 3.4-3.7% by years four and five, contradicting an immediate event-year jump with no later appreciation.","results_keys":["reference_checks"],"status":"supported","caveat":"Only the stacked and Sun-Abraham paths in Section 3.2/Table 3 were used."},4 {"id":"C3","text":"R1's attributed finding (ATT -0.431, SE 0.210, 5% significance, ~35%) matches Table 6 of arXiv:2609.27732v1, although the source's Table A19 prints SE 0.218 for the same estimate.","results_keys":["reference_checks","data_issues"],"status":"supported","caveat":"Internal source inconsistency is not resolvable from the extract; significance at 5% holds under either printed SE."},5 {"id":"C4","text":"R4's entry year 2025 is consistent because it equals the year of the first arXiv submission, even though the reproduced version (v3) is from 2026.","results_keys":["reference_checks"],"status":"supported","caveat":"Applies the year rule stated in TASK.md."},6 {"id":"C5","text":"All five extracts were analysed with no exclusions, and five data issues (rendering defects and internal inconsistencies) were recorded.","results_keys":["n_analysed","n_input_rows","data_issues"],"status":"supported","caveat":""}7]8