{"id":"b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b","title":"ScalableRAG authors propose exact corpus operations without costly ingestion","argument":"Authors Hilaf Hasson, Aditya Chakravarty, Jayant Thomas, and Krishna Gogineni introduce ScalableRAG, a system that lets an agent build and combine named document and value sets at query time. They argue this closes a gap in retrieval-augmented generation (RAG): lightweight retrievers can find evidence but do not expose the corpus-wide operations needed for counts, intersections, and aggregations, while systems that support such operations require expensive ingestion.\n\nThe authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional question-answering benchmarks. On the aggregative Hotels-50 benchmark, they report a score of 71.74, while removing the workspace lowers the score to 38.4. A second version, Limited-Ingestion ScalableRAG, adds a lightweight index; the authors report results above a judge score of 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.\n\nThe paper’s central argument is that a query-time workspace can combine exact corpus operations with low ingestion cost and scale to large collections. These are reported benchmark results, not evidence here of performance across other workloads. How well would the approach perform on real-world collections and questions beyond the reported benchmarks?","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The evidence is the paper’s abstract and reports benchmark results; it does not establish performance across real-world collections or workloads beyond those evaluated.","speaker":{"name":"Hasson, Hilaf, Chakravarty, Aditya, Thomas, Jayant, Gogineni, Krishna","roleAtSource":"Article author"},"source":{"title":"ScalableRAG: High-Quality RAG with Exact Corpus Operations at Minimal Ingestion Cost","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2607.25135","publishedAt":"2026-07-27","dateBasis":"Date basis not provided."},"evidence":[{"text":"ScalableRAG uses a persistent typed-set workspace for query-time corpus operations.","excerpt":"a persistent typed-set workspace: an agent builds named document and value sets at query time and composes them through set algebra, typed extraction, aggregation, and diagnostics.","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"},{"text":"The authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional QA benchmarks.","excerpt":"Zero-Ingestion ScalableRAG needs no preprocessing, not even a vector index. It matches the leading baseline on four conventional QA benchmarks","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"},{"text":"The authors report a Hotels-50 score of 71.74, falling to 38.4 when the workspace is removed.","excerpt":"on aggregative Hotels-50 beats every evaluated system, scoring 71.74: 36.23 above the leading baseline (SRAG), and 40.22 points above A-RAG. Removing the workspace drops Hotels-50 to 38.4.","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"},{"text":"The authors report that Limited-Ingestion ScalableRAG scores above 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.","excerpt":"is the only evaluated system above 80 judge score at 5M documents, at 13 to 16 seconds per question and under 20 minutes ingestion","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2607.25135","checkedAt":"2026-11-14T00:16:08.192Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How well would ScalableRAG perform on real-world collections and questions beyond the reported benchmarks?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"ScalableRAG authors propose exact corpus operations without costly ingestion","summary":"Authors Hilaf Hasson, Aditya Chakravarty, Jayant Thomas, and Krishna Gogineni introduce ScalableRAG, a system that lets an agent build and combine named document and value sets at query time. They argue this closes a gap in retrieval-augmented generation (RAG): lightweight retrievers can find evidence but do not expose the corpus-wide operations needed for counts, intersections, and aggregations, while systems that support such operations require expensive ingestion.\n\nThe authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional question-answering benchmarks. On the aggregative Hotels-50 benchmark, they report a score of 71.74, while removing the workspace lowers the score to 38.4. A second version, Limited-Ingestion ScalableRAG, adds a lightweight index; the authors report results above a judge score of 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.\n\nThe paper’s central argument is that a query-time workspace can combine exact corpus operations with low ingestion cost and scale to large collections. These are reported benchmark results, not evidence here of performance across other workloads. How well would the approach perform on real-world collections and questions beyond the reported benchmarks?","question":"How well would ScalableRAG perform on real-world collections and questions beyond the reported benchmarks?","claims":[{"text":"ScalableRAG uses a persistent typed-set workspace for query-time corpus operations.","excerpt":"a persistent typed-set workspace: an agent builds named document and value sets at query time and composes them through set algebra, typed extraction, aggregation, and diagnostics.","support":"source-stated"},{"text":"The authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional QA benchmarks.","excerpt":"Zero-Ingestion ScalableRAG needs no preprocessing, not even a vector index. It matches the leading baseline on four conventional QA benchmarks","support":"source-stated"},{"text":"The authors report a Hotels-50 score of 71.74, falling to 38.4 when the workspace is removed.","excerpt":"on aggregative Hotels-50 beats every evaluated system, scoring 71.74: 36.23 above the leading baseline (SRAG), and 40.22 points above A-RAG. Removing the workspace drops Hotels-50 to 38.4.","support":"source-stated"},{"text":"The authors report that Limited-Ingestion ScalableRAG scores above 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.","excerpt":"is the only evaluated system above 80 judge score at 5M documents, at 13 to 16 seconds per question and under 20 minutes ingestion","support":"source-stated"}],"limitations":["The evidence is the paper’s abstract and reports benchmark results; it does not establish performance across real-world collections or workloads beyond those evaluated."],"mediaId":"7d777c1151f09b21e6403caf0da5409ade7d4b2d95f65636b453401ab4964654","correctionReason":null,"id":"b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b","publicationId":"bbbbae51-e1f7-4b10-8fa9-99d50e8d74f0","version":1,"sourceUrl":"https://arxiv.org/abs/2607.25135","sourceTitle":"ScalableRAG: High-Quality RAG with Exact Corpus Operations at Minimal Ingestion Cost","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-07-27","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-10-28T05:56:36.597Z","createdAt":"2026-11-14T00:01:34.152Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"7d777c1151f09b21e6403caf0da5409ade7d4b2d95f65636b453401ab4964654","kind":"article","url":"https://arxiv.org/abs/2607.25135","sourceUrl":"https://arxiv.org/abs/2607.25135","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-10-30T20:17:17.320Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2607.25135","checkedAt":"2026-11-14T00:16:08.192Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"7d777c1151f09b21e6403caf0da5409ade7d4b2d95f65636b453401ab4964654","confidence":0.85,"provider":"typesafe","model":"jev-1.13.0","inputHash":"f9fefc00568ed5b509227c4f60428dde99374b0ac397642ef687f95fd3dfbd6a","decidedAt":"2026-10-30T20:17:17.771Z"},"provenance":{"textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observationId":"7470ca99ed14e029704f2664fb16af59dd38679b3a5098de92d55a434562f465","retrievedAt":"2026-10-30T20:17:17.249Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"b7add350-cdb9-4ecb-8e3f-b81e1c55bf5c"},"reviewer":{"model":"gpt-6-luna","runId":"488f8b61-6a1f-49c1-80c8-c1ca08c413ed"}},"audit":{"status":"passed","reviewedAt":"2026-11-14T00:01:34.152Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Hasson, Hilaf","url":null,"kind":"unknown"},{"name":"Chakravarty, Aditya","url":null,"kind":"unknown"},{"name":"Thomas, Jayant","url":null,"kind":"unknown"},{"name":"Gogineni, Krishna","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Hasson, Hilaf","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Hasson, Hilaf","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}},{"actor":{"name":"Chakravarty, Aditya","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Chakravarty, Aditya","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}},{"actor":{"name":"Thomas, Jayant","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Thomas, Jayant","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}},{"actor":{"name":"Gogineni, Krishna","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Gogineni, Krishna","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2607.25135","snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z","basis":"publisher-page-metadata","sourceLinks":46,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b/source"}},"canonicalUrl":"https://escapemeridian.com/readings/scalablerag-authors-propose-exact-corpus-operations-without-b5340e0f","formats":{"html":"/readings/scalablerag-authors-propose-exact-corpus-operations-without-b5340e0f","markdown":"/readings/scalablerag-authors-propose-exact-corpus-operations-without-b5340e0f.md","json":"/readings/scalablerag-authors-propose-exact-corpus-operations-without-b5340e0f.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"conditions":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[{"id":"term-rag","name":"Retrieval-augmented generation","explanation":"A setup in which a model first looks up relevant documents and then answers using them.","referenceUrl":"https://developers.google.com/machine-learning/glossary#retrieval-augmented-generation-rag"},{"id":"term-benchmark","name":"Benchmark","explanation":"A defined set of tests used to compare systems. Results depend on the tasks and measures selected.","referenceUrl":"https://arxiv.org/abs/2211.09110"}],"entities":[{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"},{"id":"time","name":"TIME","kind":"organization","context":"TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.","referenceUrls":["https://time.com/"],"path":"/entities/time","json":"/entities/time.json","markdown":"/entities/time.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b","revision":"bbbbae51-e1f7-4b10-8fa9-99d50e8d74f0","revisionKind":"publication-id","title":"ScalableRAG authors propose exact corpus operations without costly ingestion","url":"https://escapemeridian.com/readings/scalablerag-authors-propose-exact-corpus-operations-without-b5340e0f"},"source":{"type":"source","id":"https://arxiv.org/abs/2607.25135","revision":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","revisionKind":"sha256-source-text","title":"ScalableRAG: High-Quality RAG with Exact Corpus Operations at Minimal Ingestion Cost","url":"https://arxiv.org/abs/2607.25135"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b","title":"ScalableRAG authors propose exact corpus operations without costly ingestion","argument":"Authors Hilaf Hasson, Aditya Chakravarty, Jayant Thomas, and Krishna Gogineni introduce ScalableRAG, a system that lets an agent build and combine named document and value sets at query time. They argue this closes a gap in retrieval-augmented generation (RAG): lightweight retrievers can find evidence but do not expose the corpus-wide operations needed for counts, intersections, and aggregations, while systems that support such operations require expensive ingestion.\n\nThe authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional question-answering benchmarks. On the aggregative Hotels-50 benchmark, they report a score of 71.74, while removing the workspace lowers the score to 38.4. A second version, Limited-Ingestion ScalableRAG, adds a lightweight index; the authors report results above a judge score of 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.\n\nThe paper’s central argument is that a query-time workspace can combine exact corpus operations with low ingestion cost and scale to large collections. These are reported benchmark results, not evidence here of performance across other workloads. How well would the approach perform on real-world collections and questions beyond the reported benchmarks?","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The evidence is the paper’s abstract and reports benchmark results; it does not establish performance across real-world collections or workloads beyond those evaluated.","speaker":{"name":"Hasson, Hilaf, Chakravarty, Aditya, Thomas, Jayant, Gogineni, Krishna","roleAtSource":"Article author"},"source":{"title":"ScalableRAG: High-Quality RAG with Exact Corpus Operations at Minimal Ingestion Cost","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2607.25135","publishedAt":"2026-07-27","dateBasis":"Date basis not provided."},"evidence":[{"text":"ScalableRAG uses a persistent typed-set workspace for query-time corpus operations.","excerpt":"a persistent typed-set workspace: an agent builds named document and value sets at query time and composes them through set algebra, typed extraction, aggregation, and diagnostics.","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"},{"text":"The authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional QA benchmarks.","excerpt":"Zero-Ingestion ScalableRAG needs no preprocessing, not even a vector index. It matches the leading baseline on four conventional QA benchmarks","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"},{"text":"The authors report a Hotels-50 score of 71.74, falling to 38.4 when the workspace is removed.","excerpt":"on aggregative Hotels-50 beats every evaluated system, scoring 71.74: 36.23 above the leading baseline (SRAG), and 40.22 points above A-RAG. Removing the workspace drops Hotels-50 to 38.4.","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"},{"text":"The authors report that Limited-Ingestion ScalableRAG scores above 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.","excerpt":"is the only evaluated system above 80 judge score at 5M documents, at 13 to 16 seconds per question and under 20 minutes ingestion","url":"https://arxiv.org/abs/2607.25135","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2607.25135","checkedAt":"2026-11-14T00:16:08.192Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How well would ScalableRAG perform on real-world collections and questions beyond the reported benchmarks?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"ScalableRAG authors propose exact corpus operations without costly ingestion","summary":"Authors Hilaf Hasson, Aditya Chakravarty, Jayant Thomas, and Krishna Gogineni introduce ScalableRAG, a system that lets an agent build and combine named document and value sets at query time. They argue this closes a gap in retrieval-augmented generation (RAG): lightweight retrievers can find evidence but do not expose the corpus-wide operations needed for counts, intersections, and aggregations, while systems that support such operations require expensive ingestion.\n\nThe authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional question-answering benchmarks. On the aggregative Hotels-50 benchmark, they report a score of 71.74, while removing the workspace lowers the score to 38.4. A second version, Limited-Ingestion ScalableRAG, adds a lightweight index; the authors report results above a judge score of 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.\n\nThe paper’s central argument is that a query-time workspace can combine exact corpus operations with low ingestion cost and scale to large collections. These are reported benchmark results, not evidence here of performance across other workloads. How well would the approach perform on real-world collections and questions beyond the reported benchmarks?","question":"How well would ScalableRAG perform on real-world collections and questions beyond the reported benchmarks?","claims":[{"text":"ScalableRAG uses a persistent typed-set workspace for query-time corpus operations.","excerpt":"a persistent typed-set workspace: an agent builds named document and value sets at query time and composes them through set algebra, typed extraction, aggregation, and diagnostics.","support":"source-stated"},{"text":"The authors report that Zero-Ingestion ScalableRAG needs no preprocessing or vector index and matches the leading baseline on four conventional QA benchmarks.","excerpt":"Zero-Ingestion ScalableRAG needs no preprocessing, not even a vector index. It matches the leading baseline on four conventional QA benchmarks","support":"source-stated"},{"text":"The authors report a Hotels-50 score of 71.74, falling to 38.4 when the workspace is removed.","excerpt":"on aggregative Hotels-50 beats every evaluated system, scoring 71.74: 36.23 above the leading baseline (SRAG), and 40.22 points above A-RAG. Removing the workspace drops Hotels-50 to 38.4.","support":"source-stated"},{"text":"The authors report that Limited-Ingestion ScalableRAG scores above 80 at five million documents, with 13 to 16 seconds per question and under 20 minutes of ingestion.","excerpt":"is the only evaluated system above 80 judge score at 5M documents, at 13 to 16 seconds per question and under 20 minutes ingestion","support":"source-stated"}],"limitations":["The evidence is the paper’s abstract and reports benchmark results; it does not establish performance across real-world collections or workloads beyond those evaluated."],"mediaId":"7d777c1151f09b21e6403caf0da5409ade7d4b2d95f65636b453401ab4964654","correctionReason":null,"id":"b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b","publicationId":"bbbbae51-e1f7-4b10-8fa9-99d50e8d74f0","version":1,"sourceUrl":"https://arxiv.org/abs/2607.25135","sourceTitle":"ScalableRAG: High-Quality RAG with Exact Corpus Operations at Minimal Ingestion Cost","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-07-27","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-10-28T05:56:36.597Z","createdAt":"2026-11-14T00:01:34.152Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"7d777c1151f09b21e6403caf0da5409ade7d4b2d95f65636b453401ab4964654","kind":"article","url":"https://arxiv.org/abs/2607.25135","sourceUrl":"https://arxiv.org/abs/2607.25135","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-10-30T20:17:17.320Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2607.25135","checkedAt":"2026-11-14T00:16:08.192Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"7d777c1151f09b21e6403caf0da5409ade7d4b2d95f65636b453401ab4964654","confidence":0.85,"provider":"typesafe","model":"jev-1.13.0","inputHash":"f9fefc00568ed5b509227c4f60428dde99374b0ac397642ef687f95fd3dfbd6a","decidedAt":"2026-10-30T20:17:17.771Z"},"provenance":{"textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observationId":"7470ca99ed14e029704f2664fb16af59dd38679b3a5098de92d55a434562f465","retrievedAt":"2026-10-30T20:17:17.249Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"b7add350-cdb9-4ecb-8e3f-b81e1c55bf5c"},"reviewer":{"model":"gpt-6-luna","runId":"488f8b61-6a1f-49c1-80c8-c1ca08c413ed"}},"audit":{"status":"passed","reviewedAt":"2026-11-14T00:01:34.152Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Hasson, Hilaf","url":null,"kind":"unknown"},{"name":"Chakravarty, Aditya","url":null,"kind":"unknown"},{"name":"Thomas, Jayant","url":null,"kind":"unknown"},{"name":"Gogineni, Krishna","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Hasson, Hilaf","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Hasson, Hilaf","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}},{"actor":{"name":"Chakravarty, Aditya","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Chakravarty, Aditya","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}},{"actor":{"name":"Thomas, Jayant","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Thomas, Jayant","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}},{"actor":{"name":"Gogineni, Krishna","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Gogineni, Krishna","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2607.25135","locator":"head/meta[name=author|citation_author]"},"assertionId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:16:08.192Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","sourceUrl":"https://arxiv.org/abs/2607.25135","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2607.25135","snapshotId":"c02db2b5fec59cb7e2738334b2a7597b0271f033f8d27e804820891702abaef9","contentSha256":"5ae6431a00cad9d6c2cdc1d85791abaf0b967358862e972eec09f2732b2ad4a2","textHash":"4b9754796e035522f4ac97dd125d965559952c6f8f7f23aed0bb014975521763","observedAt":"2026-11-14T00:16:08.192Z","basis":"publisher-page-metadata","sourceLinks":46,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/b5340e0f41d1be973355a0e05dcf5057f344cd7cb07d940002db4ebce402d07b/source"}}}],"audits":[{"publicationId":"bbbbae51-e1f7-4b10-8fa9-99d50e8d74f0","verdict":"passed","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading correctly names the four listed authors and consistently frames the performance figures as author-reported results, not established facts.","excerpt":"Authors:Hilaf Hasson, Aditya Chakravarty, Jayant Thomas, Krishna Gogineni"},{"dimension":"summary","status":"pass","explanation":"The title, summary and question faithfully convey the abstract's capability-cost gap argument, the query-time workspace and both system variants. Terms such as ingestion, vector index and judge score are left to tooltips, but the core idea remains understandable to a general reader: the system is cheap to set up and can count or combine across documents. The question is a fair unresolved issue rather than a leading claim.","excerpt":"ScalableRAG closes this gap with a persistent typed-set workspace: an agent builds named document and value sets at query time"},{"dimension":"date","status":"pass","explanation":"originalPublished 2026-07-27 matches the v1 submission date. The later v2 revision and the feed date are correctly not used.","excerpt":"[Submitted on 27 Jul 2026 (v1), last revised 26 Oct 2026 (this version, v2)]"},{"dimension":"support","status":"pass","explanation":"Every numeric claim matches the abstract: 71.74 on Hotels-50, 38.4 without the workspace, above 80 at 5M documents, 13 to 16 seconds per question, and under 20 minutes of ingestion. The scaling statement reflects the authors' own million-document viability claim. The limitation correctly bounds benchmark results without inventing stakeholders or impacts.","excerpt":"ScalableRAG combines minimal ingestion, conventional-QA parity, exact corpus operations, and million-document viability; no evaluated alternative provides all four."},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct arXiv abstract page for this paper, which is the most direct access to the source.","excerpt":null}],"createdAt":"2026-11-14T00:01:34.152Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}