Files
ely-eye/bench/sets/_cache/mmlongbench_pages.json
ZacharyZhang-NY 77219c930b Add third-party benchmark suite frozen at 50,789 sources / 5,308 queries
Ground truth comes only from third-party annotations (MMLongBench-Doc
evidence_page_ids, VisRAG qrels, ScreenSpot-v2 bboxes, Video-MME QA,
LongMemEval answer_session_ids, CodeSearchNet func_path+func_name) and
from document structure; no model judgement is used anywhere.

- bench/corpus/manifest.jsonl: 28 sources with url, retrieved_at, sha256,
  annotation_source, license, files[]; verified by bench.manifest_check
- bench/sets/{pdf,image,ui_screenshot,video,code,conversation}.{jsonl} plus
  corpus indexes and a manifest that freezes every file's sha256
- 2,928 answerable + 2,380 unanswerable queries over 854 holders; every
  query carries scope {holder, scenario, as_of} for transactive routing
- bench/validate_sets.py fails on missing gold, holder mismatch between
  gold and query, gold observed after as_of, and manifest digest drift
2026-10-06 12:05:43 -04:00

138 lines
5.8 KiB
JSON

{
"05-03-18-political-release.pdf": 17,
"0b85477387a9d0cc33fca0f4becaa0e5.pdf": 16,
"0e94b4197b10096b1f4c699701570fbf.pdf": 15,
"11-21-16-Updated-Post-Election-Release.pdf": 45,
"12-15-15-ISIS-and-terrorism-release-final.pdf": 28,
"2005.12872v3.pdf": 26,
"2021-Apple-Catalog.pdf": 55,
"2023.acl-long.386.pdf": 24,
"2023.findings-emnlp.248.pdf": 19,
"2024.ug.eprospectus.pdf": 27,
"2210.02442v1.pdf": 24,
"2303.05039v2.pdf": 9,
"2303.08559v2.pdf": 30,
"2305.13186v3.pdf": 27,
"2305.14160v4.pdf": 16,
"2306.05425v1.pdf": 22,
"2307.09288v2.pdf": 77,
"2309.17421v2.pdf": 166,
"2310.05634v2.pdf": 24,
"2310.07609v1.pdf": 10,
"2310.09158v1.pdf": 23,
"2311.16502v3.pdf": 117,
"2312.04350v3.pdf": 28,
"2312.09390v1.pdf": 49,
"2312.10997v5.pdf": 21,
"2401.18059v1.pdf": 23,
"2405.09818v1.pdf": 27,
"3276a5b991c49cf5f9a4af0f7d6fce67.pdf": 16,
"379f44022bb27aa53efd5d322c7b57bf.pdf": 17,
"3M_2018_10K.pdf": 160,
"52b3137455e7ca4df65021a200aef724.pdf": 112,
"698bba535087fa9a7f9009e172a7f763.pdf": 20,
"7c3f6204b3241f142f0f8eb8e1fefe7a.pdf": 15,
"8dfc21ec151fb9d3578fc32d5c4e5df9.pdf": 18,
"8e7c4cb542ad160f80fb3d795ada35d8.pdf": 17,
"91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf": 40,
"936c0e2c2e6c8e0c07c51bfaf7fd0a83.pdf": 15,
"ACTIVISIONBLIZZARD_2019_10K.pdf": 198,
"ADOBE_2015_10K.pdf": 116,
"AMAZON_2017_10K.pdf": 85,
"BESTBUY_2023_10K.pdf": 75,
"BRO-GL-MMONEY.pdf": 16,
"Bergen-Brochure-en-2022-23.pdf": 46,
"COSTCO_2021_10K.pdf": 76,
"Campaign_038_Introducing_AC_Whitepaper_v5e.pdf": 28,
"DSA-278777.pdf": 21,
"GPL-Graduate-Studies-Professional-Learning-Brochure-Jul-2021.pdf": 17,
"Guide-for-international-students-web.pdf": 76,
"ISEP_student_handbook_2020.pdf": 24,
"Independents-Report.pdf": 23,
"Macbook_air.pdf": 76,
"NETFLIX_2015_10K.pdf": 72,
"NIKE_2021_10K.pdf": 109,
"NUS-Business-School-BBA-Brochure-2024.pdf": 24,
"NUS-FASS-Graduate-Guidebook-2021-small.pdf": 31,
"NYU_graduate.pdf": 30,
"PG_20.07.30_U.S.-Views-China_final.pdf": 22,
"PG_2020.03.09_US-Germany_FINAL.pdf": 21,
"PG_2020.05.21_International-Cooperation-COVID_FINAL.pdf": 26,
"PG_2021.03.04_US-Views-on-China_FINAL.pdf": 40,
"PH_2016.06.08_Economy-Final.pdf": 23,
"PIP_Seniors-and-Tech-Use_040314.pdf": 27,
"PI_2017.10.04_Automation_FINAL.pdf": 61,
"PI_2018.11.19_algorithms_FINAL.pdf": 30,
"PP_2019.01.17_Trump-economy_FINAL2.pdf": 28,
"PP_2020.08.06_COVID-19-Restrictions_FINAL-1.pdf": 26,
"PP_2021.04.22_voting-access_REPORT.pdf": 24,
"PRE_2022.09.29_NSL-politics_REPORT.pdf": 108,
"PS_2018.01.09_STEM_FINAL.pdf": 105,
"PWC_opportunity_of_lifetime.pdf": 18,
"Pew-Research-Center_Hispanic-Identity-Report_12.20.2017.pdf": 32,
"RAR.pdf": 28,
"SAO-StudentSupport_Guidebook-Content.pdf": 44,
"STEPBACK.pdf": 36,
"Sinopolis-Chengdu.pdf": 96,
"SnapNTell.pdf": 20,
"StudentSupport_Guidebook.pdf": 44,
"User_Manual_1500S_Classic_EN.pdf": 108,
"a4f3ced0696009fec3179f493e4f28c4.pdf": 17,
"a5879805d70c854ea4361e43a84e3bb2.pdf": 15,
"afe620b9beac86c1027b96d31d396407.pdf": 20,
"amb-siteaudits-ds15-150204174043-conversion-gate01_95.pdf": 41,
"asdaaburson-marstellerarabyouthsurvey2014-140407100615-phpapp01_95.pdf": 45,
"avalaunchpresentationsthatkickasteriskv3copy-150318114804-conversion-gate01_95.pdf": 27,
"b3m5kaeqm2w8n4bwcesw-140602121350-phpapp02_95.pdf": 21,
"bariumswallowpresentation-090810084400-phpapp01_95.pdf": 77,
"bdf54dxa.pdf": 54,
"bigdatatrends-120723191058-phpapp02_95.pdf": 53,
"c31e6580d0175ab3f9d99d1ff0bfa000.pdf": 20,
"caltraincapacitymountainview1-150701205750-lva1-app6891_95.pdf": 31,
"camry_ebrochure.pdf": 26,
"catvsdogdlpycon15se-150512122612-lva1-app6891_95.pdf": 68,
"chapter8-geneticscompatibilitymode-141214140247-conversion-gate02_95.pdf": 78,
"competitiveoutcomes-091006065143-phpapp01_95.pdf": 48,
"csewt7zsecmmbzjufbyx-signature-24d91a254426c21c3079384270e1f138dc43a271cfe15d6d520d68205855b2a3-poli-150306115347-conversion-gate01_95.pdf": 65,
"ddoseattle-150627210357-lva1-app6891_95.pdf": 63,
"digitalmeasurementframework22feb2011v6novideo-110221233835-phpapp01_95.pdf": 196,
"disciplined-agile-business-analysis-160218012713_95.pdf": 41,
"dr-vorapptchapter1emissionsources-121120210508-phpapp02_95.pdf": 196,
"e639029d16094ea71d964e2fb953952b.pdf": 20,
"e79deb02a0c0e87511080836c5d4347b.pdf": 17,
"earlybird-110722143746-phpapp02_95.pdf": 41,
"earthlinkweb-150213112111-conversion-gate02_95.pdf": 44,
"ecommerceopportunityindia-141124010546-conversion-gate01_95.pdf": 28,
"edb88a99670417f64a6b719646aed326.pdf": 20,
"efd88e41c5f2606c57929cac6c1c0605.pdf": 20,
"efis-140411041451-phpapp01_95.pdf": 32,
"f1f5242528411b262be447e61e2eb10f.pdf": 20,
"f86d073b0d735ac873a65d906ba82758.pdf": 20,
"f8d3a162ab9507e021d83dd109118b60.pdf": 17,
"fd76bbefe469561966e5387aa709c482.pdf": 17,
"fdac8d1e9ef56519371df7e6532df27d.pdf": 20,
"finalmediafindingspdf-141228031149-conversion-gate02_95.pdf": 44,
"finalpresentationdeck-whatwhyhowofcertificationsocial-160324220748_95.pdf": 35,
"formwork-150318073913-conversion-gate01_95.pdf": 60,
"germanwingsdigitalcrisisanalysis-150403064828-conversion-gate01_95.pdf": 23,
"guojixueshengshenghuozhinanyingwen9.1.pdf": 40,
"honor_watch_gs_pro.pdf": 42,
"indonesiamobilemarketresearch-ag-150106055934-conversion-gate02_95.pdf": 38,
"measuringsuccessonfacebooktwitterlinkedin-160317142140_95.pdf": 25,
"mi_phone.pdf": 30,
"mmdetection-readthedocs-io-en-v2.18.0.pdf": 468,
"nielsen2015musicbizpresentation-final-150526143534-lva1-app6891_95.pdf": 36,
"nova_y70.pdf": 45,
"obs-productdesc-en.pdf": 65,
"owners-manual-2170416.pdf": 32,
"q1-2023-bilibili-inc-investor-presentation.pdf": 27,
"reportq32015-151009093138-lva1-app6891_95.pdf": 34,
"san-francisco-11-contents.pdf": 40,
"stereo_headset.pdf": 151,
"t480_ug_en.pdf": 170,
"tacl_a_00660.pdf": 23,
"transform-software-delivery-with-valueedge-brochure.pdf": 20,
"watch_d.pdf": 27,
"welcome-to-nus.pdf": 24
}