docs(audit): DQ-3 content fidelity verified clean — assessment axis complete

Re-ran chunk_text + is_quality_chunk on all 29 source .txt files vs stored
chunks. Coverage 109-114% everywhere (the >100% is the 64/512-word overlap, i.e.
full retention); stored==kept for every paper. Only 18 chunks dropped corpus-wide,
all low_alpha = garbled OCR tables in 2 scanned-book sources (0-387-29555-0_13,
depositonce-5415) — no real math content lost. Head/tail aligned (no truncation).

Closes the data-quality assessment axis: DQ-1 (citations), DQ-2 (metadata),
DQ-3 (fidelity), DQ-4 (retrieval) all resolved. DQ-5 (re-ingest idempotency)
remains, tracked separately.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

docs/audit/DATA-QUALITY-2026-06-15.md

@@ -4,8 +4,10 @@
 applied to live DB + wired into ingest). **DQ-4 MEASURED 2026-06-16** — retrieval is
 strong after fixing a P0 `codex search paper` crash. **DQ-2 RESOLVED 2026-06-16**
 (1911.00966 fully recovered from arXiv; s00454 abstract from S2; book chapter is a
-documented limit; recovery wired into ingest). **DQ-3 still open. New: DQ-5**
-(`ingest_paper` not idempotent for DOI papers — blocks R-A/R-C).
+documented limit; recovery wired into ingest). **DQ-3 RESOLVED 2026-06-16**
+(content fidelity clean — full coverage, only OCR junk dropped). **All four
+assessment items (DQ-1..DQ-4) DONE. New: DQ-5** (`ingest_paper` not idempotent
+for DOI papers — blocks R-A/R-C, tracked separately).
 **Roadmap R-A..R-E** added 2026-06-16 (free-source acquisition levers — see Roadmap section).
 **Author:** Audit-Loop (Opus). **Intended reader:** a *cold* session with no memory
 of the audit conversation — everything needed to continue is in this file.
@@ -157,7 +159,29 @@ incomplete, or unrepresentative data. This document is that second axis — a
 - **Acceptance:** every paper has at least a bibkey + non-zero abstract embedding,
   or the exceptions are documented with rationale.
 
-### DQ-3 — Content fidelity (chunks vs source `.txt`): NOT YET VERIFIED · **MED**
+### DQ-3 — Content fidelity (chunks vs source `.txt`): VERIFIED CLEAN · **RESOLVED 2026-06-16**
+
+**Resolution (read-only; re-ran `chunk_text` + `is_quality_chunk` on all 29
+source files and compared to stored chunks):**
+- **No paper is silently gutted.** Coverage (stored chunk chars / source chars)
+  is **109–114% for every paper** — the >100% is exactly the 64/512-word overlap
+  inflation, i.e. full retention. `stored == kept` for all 29, so the live DB
+  matches what the current pipeline produces.
+- **Only 18 chunks dropped corpus-wide, all `low_alpha`** (0 short, 0 bib),
+  concentrated in `10.1007/0-387-29555-0_13` (6) and `depositonce-5415` (10).
+  Inspected: they are **garbled OCR tables** from scanned-book sources (alpha
+  0.30–0.36, e.g. `"(u: 1) 7 < u D:= ud3424 — else H®: out.v. Family 1…"`) —
+  exactly the artefacts the alpha-ratio gate targets. **No real math content lost**
+  (the handoff's feared failure mode did not occur; math prose stays > 0.40 alpha).
+- **Head/tail alignment exact** on sampled papers: chunk 0 begins at the source
+  head (title/authors/abstract captured), last chunk ends at the source tail → no
+  front/back truncation.
+- **Note:** 0 bibliography drops across the corpus → the upstream `.txt` extraction
+  had already stripped reference lists, which is why DQ-1's citation graph relies
+  on the API/GROBID sources rather than text-mined refs.
+- **Acceptance met** for all 29 papers (not just a sample). No action needed.
+
+**Original open question (for context):**
 - **Open question:** does the stored chunk set faithfully reconstruct each source
   file, or did the chunker / `filter_chunks` silently drop material (e.g. an
   abstract, a section, math-heavy passages)? The whole corpus was ingested from
@@ -324,9 +348,11 @@ is self-contained so a cold session can pick it up.
    verified strong; surfaced the zero-vector pollution that motivates DQ-2).
 3. ~~**DQ-2**~~ — **DONE 2026-06-16** (1911.00966 fully recovered; s00454 abstract
    from S2; book chapter documented; recovery wired into ingest).
-4. **DQ-5** (NEXT — `ingest_paper` not idempotent for DOI papers; **blocks R-A/R-C**;
-   careful — M-1 canonical-id territory), then **DQ-3** (chunk-vs-source fidelity).
-5. **Roadmap levers** (see section above): **R-A** (GROBID on arXiv PDFs — closes
+4. ~~**DQ-3**~~ — **DONE 2026-06-16** (content fidelity clean; full coverage, only
+   18 OCR-junk chunks dropped, no math lost). **Assessment axis DQ-1..DQ-4 complete.**
+5. **DQ-5** (NEXT — `ingest_paper` not idempotent for DOI papers; **blocks R-A/R-C**;
+   careful — M-1 canonical-id territory).
+6. **Roadmap levers** (see section above): **R-A** (GROBID on arXiv PDFs — closes
    the last 2 zero-citation theses; needs DQ-5 first), **R-B** (Crossref refs —
    abstract half already shipped in DQ-2), **R-C** (`.tex` ingest — serves DQ-3;
    needs DQ-5 first), **R-D** (F-15 coverage warning, quick). **R-E** (paywall)