Refresh GLiNER and GLiNER2 validation labels from reviewed candidate lane

Updates validation rows only. Train and holdout files unchanged. Validation labels remain agent-authored (owner_reviewed=false); scope and candidate-selection caveat recorded in README and split_stats.json.

README.md

@@ -29,8 +29,13 @@ configs:
 
 ## Annotation workflow
 
-Labels were produced with a locally run LLM through the passage-annotation harness, with
-human-in-the-loop passage review and adjudication under the project's data-use doctrine. The same passage-level annotation workflow was used for train, validation, and holdout.
+Labels are agent-produced under the project data-use doctrine; they are not owner rulings. The training and holdout splits are unchanged.
+
+### Validation review refresh
+
+The `val` labels in both `gliner` and `gliner2` were refreshed on 2026-09-29T17:54:34+00:00 using the `gliner_val_review` passage-review lane and shards 01–04. The review covered 1,260 passages and 2,171 candidate mentions drawn from published validation labels plus GLiNER predictions scoring at least 0.6. All decisions remain `annotated_by: agent`, `owner_reviewed: false`. The other 862 candidate-free validation passages retain their existing empty labels. Because review scope is candidate-generated and does not assess mentions missed by the candidate policy, this split is an agent-reviewed audit set, not independent full-recall gold.
+
+The review-specific source metadata and span counts are in `split_stats.json`.
 
 ## Labels
 
@@ -40,7 +45,7 @@ human-in-the-loop passage review and adjudication under the project's data-use d
 
 ## Provenance
 
-- bundle: `agent_annotations_v3` (local LLM plus passage-annotation harness with human-in-the-loop review)
+- bundle: `agent_annotations_v3` (agent-annotated; `owner_reviewed: false`)
 - lanes: origin_strat_1, origin_strat_2, origin_strat_3, umar_pads, umar_pads_500
 - carve: `agents-v3` (origin-stratified) — document-disjoint, so no
   two passages of one document straddle train and eval, and stratified by origin, so each
@@ -58,12 +63,12 @@ human-in-the-loop passage review and adjudication under the project's data-use d
 
 ## Splits
 
-21,073 passages · 16,474 spans · 12,831 documents, document-disjoint.
+21,073 passages · 16,630 spans · 12,831 documents, document-disjoint.
 
 | split | rows | spans | documents | NAMED | DESCRIPTIVE | VAGUE |
 | --- | ---: | ---: | ---: | ---: | ---: | ---: |
 | train | 16,843 | 12,891 | 10,265 | 9,723 | 3,142 | 26 |
-| val | 2,122 | 1,665 | 1,283 | 1,212 | 451 | 2 |
+| val | 2,122 | 1,821 | 1,283 | 1,275 | 538 | 8 |
 | holdout | 2,108 | 1,918 | 1,283 | 1,360 | 548 | 10 |
 
 ## Usage

split_stats.json

@@ -20,6 +20,25 @@
       "owner_reviewed": false,
       "train_val_unchanged": true,
       "source_predictions_sha256": "5e297d1e1f3571d4be1dddad86531a5a63df889f0eb7348150268614cdebdddd"
+    },
+    "validation_review_refresh": {
+      "lane": "gliner_val_review plus shards 01-04",
+      "source_dataset": "rafmacalaba/datause-agents-v3",
+      "source_dataset_revision": "fd00d59c2f09215941b46895cd5b2cdb88a7bc09",
+      "source_validation_sha256": "d780afca636fed28de875bccdf71fb0dfac61a95d525d14ca922fd5a1de21e5e",
+      "source_model": "rafmacalaba/gliner_agents_v3",
+      "source_model_revision": "987a8142f304448455f398fc723c50b4dca678a1",
+      "candidate_policy": "all gold spans + predictions score >= 0.6; unthresholded predictions remain in the main artifact",
+      "candidate_passages": 1260,
+      "candidate_mentions": 2171,
+      "effective_decision_rows": 2329,
+      "kept_spans": 1821,
+      "dropped_spans": 508,
+      "suppressed_overlap_or_duplicate_rows": 16,
+      "candidate_free_passages_retained": 862,
+      "owner_reviewed": false,
+      "train_holdout_unchanged": true,
+      "refreshed_at_utc": "2026-09-29T17:54:34+00:00"
     }
   },
   "splits": {
@@ -49,11 +68,11 @@
     },
     "val": {
       "rows": 2122,
-      "spans": 1665,
+      "spans": 1821,
       "labels": {
-        "NAMED_DATA": 1212,
-        "DESCRIPTIVE_DATA": 451,
-        "VAGUE_DATA": 2
+        "NAMED_DATA": 1275,
+        "DESCRIPTIVE_DATA": 538,
+        "VAGUE_DATA": 8
       },
       "documents": 1283,
       "origins": {