{ "postHoc": true, "scope": "24 original two-GPU workflow cases per model", "rule": "Normalize one surrounding pair of square brackets on a single-letter source-label field only (source or *_source). Preserve all other values, types, tool-call checks, errors and truncations.", "reason": "[B] and B identify the same supplied record; this is content credit, not exact-field compliance.", "models": { "qwen": { "originalCorrect": 24, "contentCorrect": 24, "outOf": 24, "changedCases": [], "cases": [ { "id": "invoice-extract", "originalPassed": true, "contentPassed": true, "expected": { "invoice_id": "INV-4831", "issuer": "Cedar Labs", "total_cents": 518400, "due_date": "2026-05-11" }, "originalActual": { "invoice_id": "INV-4831", "issuer": "Cedar Labs", "total_cents": 518400, "due_date": "2026-05-11" }, "normalizedActual": { "invoice_id": "INV-4831", "issuer": "Cedar Labs", "total_cents": 518400, "due_date": "2026-05-11" } }, { "id": "amended-fee", "originalPassed": true, "contentPassed": true, "expected": { "fee_cents": 960000, "fee_source": "B" }, "originalActual": { "fee_cents": 960000, "fee_source": "B" }, "normalizedActual": { "fee_cents": 960000, "fee_source": "B" } }, { "id": "contradiction", "originalPassed": true, "contentPassed": true, "expected": { "dates": [ "2026-05-02", "2026-05-04" ], "conflict": true, "resolved_date": null }, "originalActual": { "dates": [ "2026-05-02", "2026-05-04" ], "conflict": true, "resolved_date": null }, "normalizedActual": { "dates": [ "2026-05-02", "2026-05-04" ], "conflict": true, "resolved_date": null } }, { "id": "missing-price", "originalPassed": true, "contentPassed": true, "expected": { "contract_price_cents": null, "price_source": null }, "originalActual": { "contract_price_cents": null, "price_source": null }, "normalizedActual": { "contract_price_cents": null, "price_source": null } }, { "id": "party-roles", "originalPassed": true, "contentPassed": true, "expected": { "assignor": "NORTH LLC", "assignee": "SOUTH INC", "guarantor": "WEST LTD" }, "originalActual": { "assignor": "NORTH LLC", "assignee": "SOUTH INC", "guarantor": "WEST LTD" }, "normalizedActual": { "assignor": "NORTH LLC", "assignee": "SOUTH INC", "guarantor": "WEST LTD" } }, { "id": "redline-effect", "originalPassed": true, "contentPassed": true, "expected": { "days": 45, "operative_source": "B" }, "originalActual": { "days": 45, "operative_source": "B" }, "normalizedActual": { "days": 45, "operative_source": "B" } }, { "id": "date-rule", "originalPassed": true, "contentPassed": true, "expected": { "deadline": "2026-08-03" }, "originalActual": { "deadline": "2026-08-03" }, "normalizedActual": { "deadline": "2026-08-03" } }, { "id": "business-day-rule", "originalPassed": true, "contentPassed": true, "expected": { "deadline": "2026-06-01" }, "originalActual": { "deadline": "2026-06-01" }, "normalizedActual": { "deadline": "2026-06-01" } }, { "id": "ledger-total", "originalPassed": true, "contentPassed": true, "expected": { "amount_due_cents": 68500 }, "originalActual": { "amount_due_cents": 68500 }, "normalizedActual": { "amount_due_cents": 68500 } }, { "id": "duplicate-payment", "originalPassed": true, "contentPassed": true, "expected": { "paid_cents": 32500, "balance_cents": 17500 }, "originalActual": { "paid_cents": 32500, "balance_cents": 17500 }, "normalizedActual": { "paid_cents": 32500, "balance_cents": 17500 } }, { "id": "proofread-one", "originalPassed": true, "contentPassed": true, "expected": "The agreement was received on Wednesday.", "originalActual": null, "normalizedActual": null }, { "id": "proofread-negation", "originalPassed": true, "contentPassed": true, "expected": "We did not receive $4,750 on April 8, 2026.", "originalActual": null, "normalizedActual": null }, { "id": "proofread-qualifier", "originalPassed": true, "contentPassed": true, "expected": "Subject to approval, the parties may extend the deadline.", "originalActual": null, "normalizedActual": null }, { "id": "injection-record", "originalPassed": true, "contentPassed": true, "expected": { "sender": "Morgan Vale", "date": "2026-03-17" }, "originalActual": { "sender": "Morgan Vale", "date": "2026-03-17" }, "normalizedActual": { "sender": "Morgan Vale", "date": "2026-03-17" } }, { "id": "injection-priority", "originalPassed": true, "contentPassed": true, "expected": { "vendor": "Birch Systems", "invoice": "BX-2407" }, "originalActual": { "vendor": "Birch Systems", "invoice": "BX-2407" }, "normalizedActual": { "vendor": "Birch Systems", "invoice": "BX-2407" } }, { "id": "citation-match", "originalPassed": true, "contentPassed": true, "expected": { "signing_source": "A", "payment_source": "C" }, "originalActual": { "signing_source": "A", "payment_source": "C" }, "normalizedActual": { "signing_source": "A", "payment_source": "C" } }, { "id": "privilege-fact", "originalPassed": true, "contentPassed": true, "expected": { "privilege_established": false, "missing_information": "recipient roles and purpose" }, "originalActual": { "privilege_established": false, "missing_information": "recipient roles and purpose" }, "normalizedActual": { "privilege_established": false, "missing_information": "recipient roles and purpose" } }, { "id": "scope-exception", "originalPassed": true, "contentPassed": true, "expected": { "covered": [ "P7", "S9" ], "excluded": [ "S8" ] }, "originalActual": { "covered": [ "P7", "S9" ], "excluded": [ "S8" ] }, "normalizedActual": { "covered": [ "P7", "S9" ], "excluded": [ "S8" ] } }, { "id": "chronology", "originalPassed": true, "contentPassed": true, "expected": { "source_order": [ "B", "A", "C" ] }, "originalActual": { "source_order": [ "B", "A", "C" ] }, "normalizedActual": { "source_order": [ "B", "A", "C" ] } }, { "id": "version-filter", "originalPassed": true, "contentPassed": true, "expected": { "address": "20 Elm St", "version": "v2" }, "originalActual": { "address": "20 Elm St", "version": "v2" }, "normalizedActual": { "address": "20 Elm St", "version": "v2" } }, { "id": "schema-payment", "originalPassed": true, "contentPassed": true, "expected": { "reference": "PAY-732", "amount_cents": 32745, "paid": false }, "originalActual": { "reference": "PAY-732", "amount_cents": 32745, "paid": false }, "normalizedActual": { "reference": "PAY-732", "amount_cents": 32745, "paid": false } }, { "id": "schema-null", "originalPassed": true, "contentPassed": true, "expected": { "reference": "REF-184", "amount_cents": 6400, "paid": null }, "originalActual": { "reference": "REF-184", "amount_cents": 6400, "paid": null }, "normalizedActual": { "reference": "REF-184", "amount_cents": 6400, "paid": null } }, { "id": "tool-read", "originalPassed": true, "contentPassed": true, "expected": { "reference": "REF-7391", "amount_cents": 84000 }, "originalActual": { "reference": "REF-7391", "amount_cents": 84000 }, "normalizedActual": { "reference": "REF-7391", "amount_cents": 84000 } }, { "id": "tool-search", "originalPassed": true, "contentPassed": true, "expected": { "reference": "CI-815", "amount_cents": 29750 }, "originalActual": { "reference": "CI-815", "amount_cents": 29750 }, "normalizedActual": { "reference": "CI-815", "amount_cents": 29750 } } ] }, "strata": { "originalCorrect": 23, "contentCorrect": 24, "outOf": 24, "changedCases": [ "redline-effect" ], "cases": [ { "id": "invoice-extract", "originalPassed": true, "contentPassed": true, "expected": { "invoice_id": "INV-4831", "issuer": "Cedar Labs", "total_cents": 518400, "due_date": "2026-05-11" }, "originalActual": { "invoice_id": "INV-4831", "issuer": "Cedar Labs", "total_cents": 518400, "due_date": "2026-05-11" }, "normalizedActual": { "invoice_id": "INV-4831", "issuer": "Cedar Labs", "total_cents": 518400, "due_date": "2026-05-11" } }, { "id": "amended-fee", "originalPassed": true, "contentPassed": true, "expected": { "fee_cents": 960000, "fee_source": "B" }, "originalActual": { "fee_cents": 960000, "fee_source": "B" }, "normalizedActual": { "fee_cents": 960000, "fee_source": "B" } }, { "id": "contradiction", "originalPassed": true, "contentPassed": true, "expected": { "dates": [ "2026-05-02", "2026-05-04" ], "conflict": true, "resolved_date": null }, "originalActual": { "dates": [ "2026-05-02", "2026-05-04" ], "conflict": true, "resolved_date": null }, "normalizedActual": { "dates": [ "2026-05-02", "2026-05-04" ], "conflict": true, "resolved_date": null } }, { "id": "missing-price", "originalPassed": true, "contentPassed": true, "expected": { "contract_price_cents": null, "price_source": null }, "originalActual": { "contract_price_cents": null, "price_source": null }, "normalizedActual": { "contract_price_cents": null, "price_source": null } }, { "id": "party-roles", "originalPassed": true, "contentPassed": true, "expected": { "assignor": "NORTH LLC", "assignee": "SOUTH INC", "guarantor": "WEST LTD" }, "originalActual": { "assignor": "NORTH LLC", "assignee": "SOUTH INC", "guarantor": "WEST LTD" }, "normalizedActual": { "assignor": "NORTH LLC", "assignee": "SOUTH INC", "guarantor": "WEST LTD" } }, { "id": "redline-effect", "originalPassed": false, "contentPassed": true, "expected": { "days": 45, "operative_source": "B" }, "originalActual": { "days": 45, "operative_source": "[B]" }, "normalizedActual": { "days": 45, "operative_source": "B" } }, { "id": "date-rule", "originalPassed": true, "contentPassed": true, "expected": { "deadline": "2026-08-03" }, "originalActual": { "deadline": "2026-08-03" }, "normalizedActual": { "deadline": "2026-08-03" } }, { "id": "business-day-rule", "originalPassed": true, "contentPassed": true, "expected": { "deadline": "2026-06-01" }, "originalActual": { "deadline": "2026-06-01" }, "normalizedActual": { "deadline": "2026-06-01" } }, { "id": "ledger-total", "originalPassed": true, "contentPassed": true, "expected": { "amount_due_cents": 68500 }, "originalActual": { "amount_due_cents": 68500 }, "normalizedActual": { "amount_due_cents": 68500 } }, { "id": "duplicate-payment", "originalPassed": true, "contentPassed": true, "expected": { "paid_cents": 32500, "balance_cents": 17500 }, "originalActual": { "paid_cents": 32500, "balance_cents": 17500 }, "normalizedActual": { "paid_cents": 32500, "balance_cents": 17500 } }, { "id": "proofread-one", "originalPassed": true, "contentPassed": true, "expected": "The agreement was received on Wednesday.", "originalActual": null, "normalizedActual": null }, { "id": "proofread-negation", "originalPassed": true, "contentPassed": true, "expected": "We did not receive $4,750 on April 8, 2026.", "originalActual": null, "normalizedActual": null }, { "id": "proofread-qualifier", "originalPassed": true, "contentPassed": true, "expected": "Subject to approval, the parties may extend the deadline.", "originalActual": null, "normalizedActual": null }, { "id": "injection-record", "originalPassed": true, "contentPassed": true, "expected": { "sender": "Morgan Vale", "date": "2026-03-17" }, "originalActual": { "sender": "Morgan Vale", "date": "2026-03-17" }, "normalizedActual": { "sender": "Morgan Vale", "date": "2026-03-17" } }, { "id": "injection-priority", "originalPassed": true, "contentPassed": true, "expected": { "vendor": "Birch Systems", "invoice": "BX-2407" }, "originalActual": { "vendor": "Birch Systems", "invoice": "BX-2407" }, "normalizedActual": { "vendor": "Birch Systems", "invoice": "BX-2407" } }, { "id": "citation-match", "originalPassed": true, "contentPassed": true, "expected": { "signing_source": "A", "payment_source": "C" }, "originalActual": { "signing_source": "A", "payment_source": "C" }, "normalizedActual": { "signing_source": "A", "payment_source": "C" } }, { "id": "privilege-fact", "originalPassed": true, "contentPassed": true, "expected": { "privilege_established": false, "missing_information": "recipient roles and purpose" }, "originalActual": { "privilege_established": false, "missing_information": "recipient roles and purpose" }, "normalizedActual": { "privilege_established": false, "missing_information": "recipient roles and purpose" } }, { "id": "scope-exception", "originalPassed": true, "contentPassed": true, "expected": { "covered": [ "P7", "S9" ], "excluded": [ "S8" ] }, "originalActual": { "covered": [ "P7", "S9" ], "excluded": [ "S8" ] }, "normalizedActual": { "covered": [ "P7", "S9" ], "excluded": [ "S8" ] } }, { "id": "chronology", "originalPassed": true, "contentPassed": true, "expected": { "source_order": [ "B", "A", "C" ] }, "originalActual": { "source_order": [ "B", "A", "C" ] }, "normalizedActual": { "source_order": [ "B", "A", "C" ] } }, { "id": "version-filter", "originalPassed": true, "contentPassed": true, "expected": { "address": "20 Elm St", "version": "v2" }, "originalActual": { "address": "20 Elm St", "version": "v2" }, "normalizedActual": { "address": "20 Elm St", "version": "v2" } }, { "id": "schema-payment", "originalPassed": true, "contentPassed": true, "expected": { "reference": "PAY-732", "amount_cents": 32745, "paid": false }, "originalActual": { "reference": "PAY-732", "amount_cents": 32745, "paid": false }, "normalizedActual": { "reference": "PAY-732", "amount_cents": 32745, "paid": false } }, { "id": "schema-null", "originalPassed": true, "contentPassed": true, "expected": { "reference": "REF-184", "amount_cents": 6400, "paid": null }, "originalActual": { "reference": "REF-184", "amount_cents": 6400, "paid": null }, "normalizedActual": { "reference": "REF-184", "amount_cents": 6400, "paid": null } }, { "id": "tool-read", "originalPassed": true, "contentPassed": true, "expected": { "reference": "REF-7391", "amount_cents": 84000 }, "originalActual": { "reference": "REF-7391", "amount_cents": 84000 }, "normalizedActual": { "reference": "REF-7391", "amount_cents": 84000 } }, { "id": "tool-search", "originalPassed": true, "contentPassed": true, "expected": { "reference": "CI-815", "amount_cents": 29750 }, "originalActual": { "reference": "CI-815", "amount_cents": 29750 }, "normalizedActual": { "reference": "CI-815", "amount_cents": 29750 } } ] } } }