Upload README.md with huggingface_hub

README.md

@@ -30,6 +30,12 @@ configs:
 generated with Gemma 3 27B over a merged Romanian corpus and mined for hard
 negatives with BM25.
 
+A late-interaction variant of the same triplets is at
+[`PaulBurca2005/ro-retrieval-triplets-late-interaction`](https://huggingface.co/datasets/PaulBurca2005/ro-retrieval-triplets-late-interaction).
+The two are close to disjoint - over the queries both contain, mean Jaccard
+overlap between their negative sets is 0.053 and 73% share no negative at
+all - so training on the union exposes a model to both failure modes.
+
 The column layout follows
 [`alina0195/ro-msmarco-divided`](https://huggingface.co/datasets/alina0195/ro-msmarco-divided),
 with one column added.
@@ -38,7 +44,7 @@ with one column added.
 |---|---|
 | `anchor` | the query |
 | `positive` | the document the query was generated from |
-| `negative` | one BM25-mined hard negative |
+| `negative` | one hard negative, mined by BM25 |
 | `query_source` | the upstream dataset or outlet the query comes from |
 
 One row per (query, positive, negative) pair, so a query with four negatives is
@@ -50,13 +56,15 @@ The positive is not retrieved, it is known: every query was generated *from* a
 document, so that document is the positive by construction. Retrieval is used
 only to mine negatives.
 
-Candidates come from the BM25 top-100 within the positive's own document type.
-A candidate becomes a negative only if it scores in a band relative to the
-positive's own score for that query, shares enough of the query's idf mass,
-contains one of the query's rarest terms, and is not a near-copy of the positive
-or of a negative already picked. The corpus is deduplicated by a near-duplicate
-key first, boilerplate is dropped, and self-referential queries ("what does the
-article say…") are removed.
+Candidates come from the BM25 top-100 within the positive's own
+document type. A candidate becomes a negative only if it scores in a band
+relative to the positive's own score for that query, shares enough of the
+query's idf mass, contains one of the query's rarest terms, and is not a
+near-copy of the positive or of a negative already picked.
+
+The corpus is deduplicated by a near-duplicate key first, boilerplate is
+dropped, and self-referential queries ("what does the article say…") are
+removed - identically in both variants, so the two are comparable.
 
 ## Sources