| { |
| "@context": { |
| "@language": "en", |
| "@vocab": "https://schema.org/", |
| "sc": "https://schema.org/", |
| "cr": "http://mlcommons.org/croissant/", |
| "rai": "http://mlcommons.org/croissant/RAI/", |
| "dct": "http://purl.org/dc/terms/", |
| "citeAs": "cr:citeAs", |
| "conformsTo": "dct:conformsTo", |
| "data": { |
| "@id": "cr:data", |
| "@type": "@json" |
| }, |
| "dataCollection": "rai:dataCollection", |
| "dataCollectionRawData": "rai:dataCollectionRawData", |
| "dataCollectionTimeframe": "rai:dataCollectionTimeframe", |
| "dataCollectionType": "rai:dataCollectionType", |
| "dataPreprocessingProtocol": "rai:dataPreprocessingProtocol", |
| "dataReleaseMaintenancePlan": "rai:dataReleaseMaintenancePlan", |
| "dataUseCases": "rai:dataUseCases", |
| "dataLimitations": "rai:dataLimitations", |
| "dataBiases": "rai:dataBiases", |
| "personalSensitiveInformation": "rai:personalSensitiveInformation", |
| "dataSocialImpact": "rai:dataSocialImpact", |
| "dataAnnotationProtocol": "rai:dataAnnotationProtocol", |
| "dataAnnotationPlatform": "rai:dataAnnotationPlatform", |
| "dataAnnotationAnalysis": "rai:dataAnnotationAnalysis", |
| "examples": { |
| "@id": "cr:examples", |
| "@type": "@json" |
| }, |
| "extract": "cr:extract", |
| "field": "cr:field", |
| "fileObject": "cr:fileObject", |
| "fileProperty": "cr:fileProperty", |
| "fileSet": "cr:fileSet", |
| "format": "cr:format", |
| "includes": "cr:includes", |
| "jsonPath": "cr:jsonPath", |
| "path": "cr:path", |
| "recordSet": "cr:recordSet", |
| "source": "cr:source" |
| }, |
| "@type": "sc:Dataset", |
| "name": "SFD Archive 1B-Token Source-Format Sample", |
| "alternateName": "sfd-archive-1b-source-format-sample", |
| "description": "Archive-wide SFD source-format analysis sample containing sanitized per-filing parse statistics, sampled accession metadata, aggregate summaries, and source-format figures.", |
| "url": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample", |
| "sameAs": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample", |
| "conformsTo": [ |
| "http://mlcommons.org/croissant/1.1", |
| "http://mlcommons.org/croissant/RAI/1.0" |
| ], |
| "license": "https://creativecommons.org/licenses/by-nc/4.0/", |
| "creator": [ |
| { |
| "@type": "Organization", |
| "name": "SFD Authors" |
| } |
| ], |
| "publisher": { |
| "@type": "Organization", |
| "name": "sfd-anonymous" |
| }, |
| "datePublished": "2026-05-07", |
| "version": "1.0", |
| "keywords": [ |
| "SEC", |
| "EDGAR", |
| "SFD", |
| "source-format analysis", |
| "finance", |
| "filings" |
| ], |
| "distribution": [ |
| { |
| "@type": "cr:FileObject", |
| "@id": "repo_root", |
| "name": "Hugging Face dataset repository", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/tree/main", |
| "encodingFormat": "inode/directory" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "summary_json", |
| "name": "summary.json", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/resolve/main/summary.json", |
| "encodingFormat": "application/json", |
| "contentSize": 2922 |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "source_summary_checkpoint_json", |
| "name": "source_summary_checkpoint.json", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/resolve/main/source_summary_checkpoint.json", |
| "encodingFormat": "application/json", |
| "contentSize": 32190 |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "filing_stats_jsonl_gz", |
| "name": "filing_stats.jsonl.gz", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/resolve/main/filing_stats.jsonl.gz", |
| "encodingFormat": "application/octet-stream", |
| "contentSize": 3758901 |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "sample_manifest_jsonl_gz", |
| "name": "sample_manifest.jsonl.gz", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/resolve/main/sample_manifest.jsonl.gz", |
| "encodingFormat": "application/octet-stream", |
| "contentSize": 5522301 |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "source_format_live_csv", |
| "name": "source_format_live.csv", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/resolve/main/source_format_live.csv", |
| "encodingFormat": "text/csv", |
| "contentSize": 5945 |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "source_format_live_png", |
| "name": "source_format_live.png", |
| "contentUrl": "https://huggingface.co/datasets/sfd-anonymous/sfd-archive-1b-source-format-sample/resolve/main/source_format_live.png", |
| "encodingFormat": "image/png", |
| "contentSize": 125750 |
| } |
| ], |
| "recordSet": [ |
| { |
| "@type": "cr:RecordSet", |
| "@id": "filing_stats", |
| "name": "Per-filing parse statistics", |
| "description": "One row per attempted filing parse, with status, form metadata, token counts, source-format token counts, table estimates, and PDF page counts.", |
| "field": [ |
| { |
| "@type": "cr:Field", |
| "name": "accession", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "cik", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "form_type", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "date_filed", |
| "dataType": "sc:Date", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "status", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "final_token_count", |
| "dataType": "sc:Integer", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "source_token_counts", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "filing_stats_jsonl_gz" |
| } |
| } |
| } |
| ] |
| }, |
| { |
| "@type": "cr:RecordSet", |
| "@id": "sample_manifest", |
| "name": "Sample manifest", |
| "description": "Candidate filing manifest used to construct the source-format analysis sample.", |
| "field": [ |
| { |
| "@type": "cr:Field", |
| "name": "accession", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "sample_manifest_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "cik", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "sample_manifest_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "form_type", |
| "dataType": "sc:Text", |
| "source": { |
| "fileObject": { |
| "@id": "sample_manifest_jsonl_gz" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "name": "date_filed", |
| "dataType": "sc:Date", |
| "source": { |
| "fileObject": { |
| "@id": "sample_manifest_jsonl_gz" |
| } |
| } |
| } |
| ] |
| } |
| ], |
| "dataCollectionRawData": "Public SEC EDGAR filings and derived benchmark artifacts built from those filings.", |
| "dataUseCases": [ |
| "Financial language modeling and long-context pretraining research.", |
| "Document parsing, table extraction, OCR evaluation, and filing-grounded financial reasoning.", |
| "Dataset analysis and benchmark reproducibility for the associated NeurIPS submission." |
| ], |
| "dataLimitations": [ |
| "SEC filings are not representative of all businesses or all global financial disclosures.", |
| "The artifacts reflect parser behavior and sampling choices at construction time.", |
| "Benchmarks are intended for research evaluation, not financial, legal, or investment advice." |
| ], |
| "dataBiases": [ |
| "The archive over-represents U.S.-regulated registrants, public companies, investment funds, and entities required to file with the SEC.", |
| "Filing volume and token mass are skewed by form type, year, issuer size, and filing-agent workflows." |
| ], |
| "personalSensitiveInformation": [ |
| "Underlying SEC filings are public regulatory records and may contain names, signatures, ownership disclosures, addresses, or other business-related personal information.", |
| "Synthetic benchmark transformations replace many entities and values, but users should still treat all artifacts as derived from public regulatory filings." |
| ], |
| "dataReleaseMaintenancePlan": "Artifacts are versioned on Hugging Face. Updates, replacements, or corrections will be documented in repository commits and dataset cards.", |
| "dataSocialImpact": "The datasets are intended to improve reproducible research on factual financial documents, while requiring care because downstream uses can affect financial analysis and compliance workflows.", |
| "dataCollection": "Filings were sampled from the SEC EDGAR archive and parsed through SFD to measure source-format composition and parser statistics.", |
| "dataCollectionType": [ |
| "Document analysis", |
| "Software Collection", |
| "Secondary Data analysis" |
| ], |
| "dataCollectionTimeframe": "SEC EDGAR archive-wide sample; filing dates span multiple EDGAR generations.", |
| "dataPreprocessingProtocol": "Raw filings were downloaded, parsed into SFD MultiMarkdown/statistics, summarized by source format, and sanitized by removing parser stdout tails and local paths before release." |
| } |
|
|
