{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "khattab-zaharia-colbert-lateinteraction-2020",
    "title": "Khattab and Zaharia: contextualized token-level late-interaction retrieval",
    "summary": "Encoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search.",
    "boundary": "BERT, vector indexes and neural ranking are inherited. Multivector late interaction increases index storage; benchmark speedups do not guarantee every corpus/hardware/workload tradeoff.",
    "area": "Neural information retrieval",
    "topic": "Artificial intelligence",
    "year": 2020,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "khattab-zaharia-colbert-lateinteraction-2020",
    "target": {
      "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
      "version": 1,
      "content_hash": "sha256:97b949c9c66800dca7ad27434c0ba658a7d28f57fe6a50bde96ec582d42642f4"
    },
    "humans": [
      "Omar Khattab",
      "Matei Zaharia"
    ],
    "authors_complete": true,
    "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.",
    "agents": [],
    "related": [],
    "subresults": [],
    "scientific_assessment": {
      "source_review": "SOURCE_CURATED",
      "scientific_review": "NOT_REVIEWED_BY_WRS",
      "formal_replay": "NOT_RUN",
      "provider_attestation": "NONE"
    },
    "sources": [
      {
        "id": "cur20261015-arxiv-2004-12832v1",
        "title": "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT",
        "url": "https://arxiv.org/abs/2004.12832v1"
      }
    ],
    "points": "150000",
    "published_grade": {
      "id": "khattab-zaharia-colbert-lateinteraction-2020",
      "entry_id": "khattab-zaharia-colbert-lateinteraction-2020",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
        "version": 1,
        "content_hash": "sha256:97b949c9c66800dca7ad27434c0ba658a7d28f57fe6a50bde96ec582d42642f4"
      },
      "what_changed": "Encoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search.",
      "scope": "Encoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search. BERT, vector indexes and neural ranking are inherited. Multivector late interaction increases index storage; benchmark speedups do not guarantee every corpus/hardware/workload tradeoff.",
      "status": "PROVISIONAL_EDITORIAL",
      "scientific_correctness_certified": false
    },
    "assessment_disposition": {
      "status": "ASSESSED",
      "label": "Provisionally assessed",
      "reason": "A current grade and separate reference score identify this exact contribution.",
      "target": {
        "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
        "version": 1,
        "content_hash": "sha256:97b949c9c66800dca7ad27434c0ba658a7d28f57fe6a50bde96ec582d42642f4"
      },
      "edition": "ratings-e039106f30dec177"
    }
  },
  "version": {
    "kind": "PUBLIC_RECORD_VIEW",
    "previous": null,
    "published_at": "2026-10-16T00:00:00Z",
    "publisher": {
      "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33",
      "role": "WRS_EDITORIAL_PUBLICATION"
    },
    "record": {
      "contributors": [
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Omar Khattab",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT",
            "url": "https://arxiv.org/abs/2004.12832v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Matei Zaharia",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT",
            "url": "https://arxiv.org/abs/2004.12832v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT",
            "url": "https://arxiv.org/abs/2004.12832v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "BERT, vector indexes and neural ranking are inherited. Multivector late interaction increases index storage; benchmark speedups do not guarantee every corpus/hardware/workload tradeoff.",
        "domain": "Neural information retrieval"
      },
      "sources": [
        {
          "citation": "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT",
          "url": "https://arxiv.org/abs/2004.12832v1"
        }
      ],
      "statement": "# Khattab and Zaharia: contextualized token-level late-interaction retrieval\n\nEncoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search.\n\n## Scientific scope\n\nBERT, vector indexes and neural ranking are inherited. Multivector late interaction increases index storage; benchmark speedups do not guarantee every corpus/hardware/workload tradeoff.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nOmar Khattab; Matei Zaharia\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT](https://arxiv.org/abs/2004.12832v1)\n",
      "summary": "Encoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search.",
      "title": "Khattab and Zaharia: contextualized token-level late-interaction retrieval"
    },
    "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
      "version": 1,
      "content_hash": "sha256:97b949c9c66800dca7ad27434c0ba658a7d28f57fe6a50bde96ec582d42642f4"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:4fac73d8cf625741b1e2f3ac3568aacf93d0cda9d2427b4ab0bc494039098cce"
  },
  "target": {
    "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
    "version": 1,
    "content_hash": "sha256:97b949c9c66800dca7ad27434c0ba658a7d28f57fe6a50bde96ec582d42642f4"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "khattab-zaharia-colbert-lateinteraction-2020",
    "entry_id": "khattab-zaharia-colbert-lateinteraction-2020",
    "revision": 1,
    "title": "Khattab and Zaharia: contextualized token-level late-interaction retrieval",
    "context": "Neural information retrieval",
    "track": "Human-designed research method or resource",
    "band": "A",
    "modifier": null,
    "status": "PROVISIONAL_EDITORIAL",
    "assessed_on": "2026-10-16",
    "evidence_cutoff": "2026-10-16",
    "achievement_state": "SUPPORTED_SCOPED_RESULT",
    "what_changed": "Encoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search.",
    "scope": "Encoded queries and documents independently with BERT, retained multiple contextualized token vectors and used lightweight late interaction to support offline document encoding and scalable passage search. BERT, vector indexes and neural ranking are inherited. Multivector late interaction increases index storage; benchmark speedups do not guarantee every corpus/hardware/workload tradeoff.",
    "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. BERT, vector indexes and neural ranking are inherited. Multivector late interaction increases index storage; benchmark speedups do not guarantee every corpus/hardware/workload tradeoff.",
    "references": [
      {
        "title": "ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT",
        "url": "https://arxiv.org/abs/2004.12832v1",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      }
    ],
    "roles": [
      {
        "who": "Source-listed authors and technical contributors",
        "credit": "SCIENCE",
        "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities."
      }
    ],
    "target": {
      "record_id": "urn:uuid:a9b8a4ff-d485-4cff-ab3f-5d72f14771d3",
      "version": 1,
      "content_hash": "sha256:97b949c9c66800dca7ad27434c0ba658a7d28f57fe6a50bde96ec582d42642f4"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "150000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}