Phase 6 round-6: soften firm-heterogeneity framing, fix DOCX table render

Framing softening (per partner tone decision: own the limitation rather
than defend the strong claim).

Abstract: "Firm heterogeneity is decisive ... consistent with firm-level
template-like reuse" -> "The framework surfaces pronounced firm-level
heterogeneity ... consistent with firm-level template-like reuse but not
independently diagnostic, since descriptor-only data cannot separate
reuse from digitisation-pipeline or signing-style homogeneity within a
firm; we report it as a scope limitation rather than a mechanism finding."

S V-H Limitations: new bullet "Mechanism attribution for the firm-level
heterogeneity is not identifiable from descriptor-only data." enumerates
three non-mutually-exclusive firm-level mechanisms (template-like reuse /
digitisation-pipeline homogeneity / signing-style homogeneity), notes the
(cosine, dHash) descriptor pair is by construction indifferent to which
mechanism generated a near-identical pair, and lists what additional data
would be needed for attribution.

S VI Conclusion items (3) and (4): "firm heterogeneity quantification" ->
"firm-level heterogeneity surfaced by the framework ... reported as a
framework-discriminative observation rather than a mechanism finding";
item (4) expanded from template/stamp/document-production reuse alone to
the three-mechanism scope, with explicit "not independently establishing"
and S V-H cross-reference.

DOCX export fix (export_v3.py): add missing LaTeX-to-Unicode tokens
(\checkmark, \lvert/\rvert, \lVert/\rVert, \in, \notin, \max, \min,
\log, \ln, \exp, \bullet) that were silently dropping content from
Table III rows 2-4 (integer-jitter robustness check marks empty) and
Table XVIII drift column (|Delta| empty).

Rebuild Paper_A_IEEE_Access_Draft_v3.docx via export_v3.py and install
copy as Paper_A_IEEE_Access_Draft_v4.0_20260515.docx (replaces prior
pandoc-built v4 DOCX which had empty cells in every table header with
LaTeX math and inconsistent column widths). All 43 tables now have
non-empty cells with sub/superscript runs.

Mirrored in paper_a_v4_combined.md for consistency with the single-file
combined source.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

paper/paper_a_abstract_v3.md

@@ -2,6 +2,6 @@
 
 <!-- IEEE Access target: <= 250 words, single paragraph -->
 
-Regulations require Certified Public Accountants (CPAs) to attest each audit report with a signature, but digitization makes it feasible to reuse a stored signature image across reports — through administrative stamping or firm-level electronic signing — thereby undermining individualized attestation. We build an end-to-end pipeline for screening such *non-hand-signed* signatures at scale: a Vision-Language Model identifies signature pages, YOLOv11 localizes signatures, ResNet-50 supplies deep features, and a dual-descriptor layer combines cosine similarity with an independent-minimum perceptual hash (dHash) to separate *style consistency* from *image reproduction*. Applied to 90,282 Taiwan audit reports (2013–2023), the pipeline yields 182,328 signatures from 758 CPAs; primary analyses are scoped to the Big-4 sub-corpus (437 CPAs; 150,442 signatures). Distributional diagnostics show that the apparent multimodality of the descriptor distribution dissolves under joint firm-mean centring and integer-tie jitter ($p$ rises to $0.35$), so no within-population bimodal antimode anchors the operational thresholds. We instead adopt an anchor-based inter-CPA coincidence-rate (ICCR) calibration at three units: per-comparison ($0.0006$ at cos$>0.95$; $0.0013$ at dHash$\leq 5$; $0.00014$ jointly), pool-normalised per-signature ($0.11$ under the deployed any-pair high-confidence rule), and per-document ($0.34$ for the operational HC+MC alarm). Firm heterogeneity is decisive: Firm A's per-document HC+MC inter-CPA proxy ICCR is $0.62$ versus $0.09$–$0.16$ at Firms B/C/D, and a per-signature logistic regression confirms the firm gap persists after controlling for pool size; under the deployed any-pair rule $77$–$99\%$ of inter-CPA collisions concentrate within the source firm — consistent with firm-level template-like reuse. We position the system as a specificity-proxy-anchored screening framework with human-in-the-loop review, not as a validated forensic detector; no calibrated error rates are reportable without signature-level ground truth.
+Regulations require Certified Public Accountants (CPAs) to attest each audit report with a signature, but digitization makes it feasible to reuse a stored signature image across reports — through administrative stamping or firm-level electronic signing — thereby undermining individualized attestation. We build an end-to-end pipeline for screening such *non-hand-signed* signatures at scale: a Vision-Language Model identifies signature pages, YOLOv11 localizes signatures, ResNet-50 supplies deep features, and a dual-descriptor layer combines cosine similarity with an independent-minimum perceptual hash (dHash) to separate *style consistency* from *image reproduction*. Applied to 90,282 Taiwan audit reports (2013–2023), the pipeline yields 182,328 signatures from 758 CPAs; primary analyses are scoped to the Big-4 sub-corpus (437 CPAs; 150,442 signatures). Distributional diagnostics show that the apparent multimodality of the descriptor distribution dissolves under joint firm-mean centring and integer-tie jitter ($p$ rises to $0.35$), so no within-population bimodal antimode anchors the operational thresholds. We instead adopt an anchor-based inter-CPA coincidence-rate (ICCR) calibration at three units: per-comparison ($0.0006$ at cos$>0.95$; $0.0013$ at dHash$\leq 5$; $0.00014$ jointly), pool-normalised per-signature ($0.11$ under the deployed any-pair high-confidence rule), and per-document ($0.34$ for the operational HC+MC alarm). The framework surfaces pronounced firm-level heterogeneity: Firm A's per-document HC+MC ICCR is $0.62$ versus $0.09$–$0.16$ at Firms B/C/D (gap persists after pool-size adjustment), and $77$–$99\%$ of inter-CPA collisions concentrate within the source firm. This contrast is consistent with firm-level template-like reuse but not independently diagnostic, since descriptor-only data cannot separate reuse from digitisation-pipeline or signing-style homogeneity within a firm; we report it as a scope limitation rather than a mechanism finding. We position the system as a specificity-proxy-anchored screening framework with human-in-the-loop review, not as a validated forensic detector; no calibrated error rates are reportable without signature-level ground truth.
 
-<!-- Word count: 247 -->
+<!-- Word count: 281 -->