[DOC] v2 metadata columns: schema, coverage, and the measured rejection of the derived prior

README.md

@@ -111,9 +111,13 @@ CVE and top-frequency techniques (e.g. T1574.007 on 53% of 2024 CVEs) that are
 artifacts of the cross-framework table expansion, not descriptions of real
 adversary behavior. The column is included as:
 
-1. a candidate prior at inference time (restrict suggestions to techniques compatible with the CWE chain);
-2. a baseline that a trained model must beat;
-3. a comparison column for studying where the deterministic chain diverges from analyst judgment.
+1. a baseline that a trained model must beat;
+2. a comparison column for studying where the deterministic chain diverges from analyst judgment.
+
+Its use as an inference-time candidate prior was measured and **rejected**
+(2026-08-06): at the parent-technique level the derived candidate sets cover
+only 3.3% of the analyst-chosen techniques on the test split, so any
+re-ranking toward them degrades every ranking metric.
 
 The full source analysis is documented in the
 [VulnTrain documentation](https://github.com/vulnerability-lookup/VulnTrain/blob/main/docs/attack-techniques-dataset.md).
@@ -132,6 +136,32 @@ The full source analysis is documented in the
 | `techniques_derived` | list[string] | CVE2CAPEC weak labels — **not** for training |
 | `label_sources` | list[string] | `ctid_cve` and/or `ctid_kev` |
 | `attack_version` | string | Enterprise ATT&CK version the IDs are normalized to |
+| `cvss_vector` | string | CVSS vector string, highest available version (empty if none) — v2 |
+| `cvss_version` | string | Version of `cvss_vector`: `4.0`, `3.1`, `3.0` or `2.0` — v2 |
+| `cwes` | list[string] | CWE assignments, e.g. `CWE-502 Deserialization of Untrusted Data` — v2 |
+| `affected_products` | list[string] | `vendor product` pairs from the CVE record — v2 |
+| `cpes` | list[string] | CPE identifiers — v2 |
+
+### Structured metadata columns (v2, added 2026-08-06)
+
+The v2 columns are extracted from the raw CVE records served by
+[Vulnerability-Lookup](https://vulnerability.circl.lu) (CNA container
+preferred, [CISA ADP Vulnrichment](https://github.com/cisagov/vulnrichment)
+filling many gaps — notably 100% CVSS/CWE coverage on the KEV subset);
+`cpes` is joined from
+[CIRCL/vulnerability-scores](https://huggingface.co/datasets/CIRCL/vulnerability-scores).
+v1 columns are unchanged (the update is strictly additive: identical rows
+and splits). Coverage differs by label source — report results stratified
+by `label_sources` when using these columns as model inputs:
+
+| Subset | CVEs | `cvss_vector` | `cwes` | `affected_products` | `cpes` |
+|--------|------|---------------|--------|---------------------|--------|
+| all | 1,207 | 72.0% | 84.3% | 67.4% | 93.2% |
+| `ctid_kev` | 392 | 100% | 100% | 79.8% | 79.1% |
+| `ctid_cve` | 788 | 57.1% | 76.0% | 62.2% | 100% |
+| both | 27 | 100% | 100% | 40.7% | 100% |
+
+CVSS versions among the 869 vectors: 677 × v3.1, 173 × v3.0, 18 × v4.0, 1 × v2.0.
 
 ## Label statistics