Paper A v13 rev9.2: slim abstract (rev9.1 review R7)

Trim the single-paragraph abstract 436 -> 358 words (~18%) and roughly halve the
closing caveat, per the reviewer's optional R7. All headline numbers preserved
(86,071 reports, 150,442 signatures, unimodality p = 0.35, 1.2%/17.5% chance rates,
82% vs 24-35% benchmark, 262 byte-identical); the honest-scope sentence is kept but
condensed to "between-accountant specificity proxy — not a within-accountant error
rate, nor a bound on one; no pipeline separation; no single-signature labels."

Rebuild docx + PDF.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01K35dXhb9XEM1mnYz6SSHpU

paper/v13_build/paper_v13_filled.md

@@ -4,7 +4,7 @@
 
 ## Abstract
 
-Audit reports must carry each certifying accountant's signature as the mark of an individual act of endorsement, yet once reports are produced and stored digitally a saved image of that signature can be pasted onto many reports instead — by manual stamping or by an automated signing system — producing what we term non-hand-signed signatures. The signer is genuine; the open question is whether an act of signing occurred for each report, and at archive scale this question carries no ground-truth labels. We present a label-free screening system for it and apply it to 86,071 Taiwanese statutory audit reports (2013–2023), within which the four largest audit firms contribute 150,442 analyzable signatures. The system finds the signature page, detects each signature, extracts deep features, and computes two similarities against the same accountant's other signatures: a cosine similarity reflecting style and a perceptual-hash (dHash) distance reflecting pixel-level structure, on the logic that a consistent hand keeps style high while structure varies, whereas a reused image keeps both extreme. Because the archive has no labels and the data contain no natural gap (a unimodality test gives median p = 0.35 once firm effects and the hash's integer steps are removed), no cutoff can be learned; instead we calibrate a five-way rule by how often it fires by chance between unrelated accountants in a clean reference group (the non-Firm-A firms, 2013–2019), where the strict high-confidence rule fires on about 1.2% of reports and a looser advisory band on about 17.5%. Held out from calibration as a known-positive benchmark — one firm independently described by interviews as a stamping firm, making this a confirmatory check rather than a blinded test — that firm fires the strict rule on 82% of its own signatures against 24–35% at the others, while its cross-firm rate sits at the clean floor, so the signal is entirely within the firm; the contrast survives stratification by comparison-pool size and resampling clustered at the accountant level, and 262 byte-identical signatures are direct evidence of reuse. Operationally, the screen locates where reuse concentrates without being told where to look and confines human review to exceptions. We are deliberate about what is and is not claimed: we report a between-accountant specificity proxy, not a true error rate — the within-accountant false-positive rate the question would require is not estimable without labels, and our coincidence rate is not even a bound on it — we cannot separate signing practice from a firm's imaging pipeline, and we label no single signature. Calibrated on a large Chinese-signature corpus with script-agnostic descriptors, the rule serves as an operator-set reference point for comparable Chinese-signature pipelines.
+An audit report must carry each certifying accountant's signature as the mark of an individual act of endorsement, yet in digital workflows a saved image can instead be pasted onto many reports — by stamping or by an automated signing system — producing what we term non-hand-signed signatures. The signer is genuine; the question is whether signing occurred for each report, and at archive scale this carries no labels. We present a label-free screening system and apply it to 86,071 Taiwanese statutory audit reports (2013–2023), within which the four largest firms contribute 150,442 analyzable signatures. The system finds the signature page, detects signatures, extracts deep features, and computes two similarities to the same accountant's other signatures — a style cosine and a perceptual-hash (dHash) structural distance — on the logic that a consistent hand keeps style high while structure varies, whereas a reused image keeps both extreme. With no labels and no natural gap in the data (a unimodality test gives median p = 0.35 once firm effects and the hash's integer steps are removed), no cutoff can be learned; instead we calibrate a five-way rule by how often it fires by chance between unrelated accountants in a clean reference group (non-Firm-A firms, 2013–2019), where the strict rule fires on about 1.2% of reports and a looser advisory band on about 17.5%. Held out as a known-positive benchmark — one firm independently described by interviews as a stamping firm (a confirmatory check, not a blind test) — it fires the strict rule on 82% of its own signatures against 24–35% elsewhere, with its cross-firm rate at the clean floor, so the signal is entirely within-firm; the contrast survives pool-size stratification and accountant-level resampling, and 262 byte-identical signatures are direct evidence of reuse. The screen thus locates where reuse concentrates and confines human review to exceptions. We report a between-accountant specificity proxy — not a within-accountant error rate, nor a bound on one — and we neither separate signing practice from a firm's imaging pipeline nor label any single signature. Calibrated on a large Chinese-signature corpus with script-agnostic descriptors, it serves as an operator-set reference point for comparable pipelines.
 
 **Keywords:** signature analysis, document forensics, perceptual hashing, deep features, unsupervised calibration, audit reports, anchor-based screening.