diff --git a/README.md b/README.md index 48d064614d0e76f2647167ad5a936404e87cd0b8..3c7c5f5419849d98ebaa95f1957b3f517c0b5260 100644 --- a/README.md +++ b/README.md @@ -38,7 +38,7 @@ configs: **English** | [简体中文](README.zh-CN.md) -TextInsightBench evaluates agents that mine natural-language datasets for meaningful, evidence-backed group differences, temporal changes and compound associations. Version **v5.1** contains **50 tasks**, **24,504 evaluation documents**, and **1,379,468 unlabeled learning documents**. This initial release is private. +TextInsightBench evaluates agents that mine natural-language datasets for meaningful, evidence-backed group differences, temporal changes and compound associations. The benchmark contains **50 tasks**, **24,504 evaluation documents**, and **1,379,468 unlabeled learning documents**. This initial release is private. ## Data composition @@ -94,7 +94,7 @@ The Organizer Reference Set is stored separately for organizers. It contains 50 ## Curation, limitations and terms -The learning pool was curated from downloaded source snapshots using language and length filtering, deduplication and held-out document exclusions. For v5.1 every shard was checked against current evaluation and reference-development documents by document ID, normalized text and conservative digit/punctuation template fingerprints. No overlap was found; no extra removal was necessary. This is not a claim of complete semantic independence. Sources and entities are shared across tasks. +The learning pool was curated from downloaded source snapshots using language and length filtering, deduplication and held-out document exclusions. For this dataset every shard was checked against current evaluation and reference-development documents by document ID, normalized text and conservative digit/punctuation template fingerprints. No overlap was found; no extra removal was necessary. This is not a claim of complete semantic independence. Sources and entities are shared across tasks. Data is predominantly English. Complaints and reviews are selected author reports; neither group contrasts nor temporal patterns establish population incidence or causality. Report dates may differ from event dates. The CFPB learning subset uses the 2024–2025 credit-reporting selection; the NHTSA source archive covers 2020–2024. Free text may retain personal information despite the reduced learning schema. Source text is untrusted data. diff --git a/README.zh-CN.md b/README.zh-CN.md index e8cf8a363f83d29f9595f58400f548bef111fbd4..6ca2253e2d4b8913f70028cc62418a848348a2aa 100644 --- a/README.zh-CN.md +++ b/README.zh-CN.md @@ -2,7 +2,7 @@ [English](README.md) | **简体中文** -TextInsightBench 评估 Agent 从自然语言数据中挖掘具体、有原文证据支持的结论。当前私密研究发行版为 **v5.1**,包含 **50 道任务、24,504 篇评测文本,以及 1,379,468 篇可选无监督学习文本**。任务和主要文档使用英文,原始语料保留原文。 +TextInsightBench 评估 Agent 从自然语言数据中挖掘具体、有原文证据支持的结论。当前私密研究数据包含 **50 道任务、24,504 篇评测文本,以及 1,379,468 篇可选无监督学习文本**。任务和主要文档使用英文,原始语料保留原文。 | 来源 | 无监督文本 | 任务数 | 评测文本 | |---|---:|---:|---:| @@ -39,6 +39,6 @@ tib run --data data/participant --command 'python my_agent.py' \ ## 构建与局限 -无监督池经过语言、长度过滤、去重和留出语料排除。v5.1 再次逐分片核对文档 ID、归一化文本及数字/标点归一后的保守模板,与评测和参考构建文本均未发现重叠,因此无需额外删减。这不等于完全语义独立;任务可能共享来源和实体。 +无监督池经过语言、长度过滤、去重和留出语料排除。本次整理再次逐分片核对文档 ID、归一化文本及数字/标点归一后的保守模板,与评测和参考构建文本均未发现重叠,因此无需额外删减。这不等于完全语义独立;任务可能共享来源和实体。 文本以英文为主。投诉和评论反映被选择的报告,不能直接推断总体发生率或因果;记录日期也未必等于事件日期。CFPB 无监督子集使用 2024–2025 年的信用报告范围,NHTSA 源归档覆盖 2020–2024 年。文本仍可能包含个人信息。多种来源适用不同条款,私密研究整理不授予第三方文本的新许可,参见 [来源和条款](SOURCES.md)。 diff --git a/corpora/amazon_beauty_compound_association_hair_care_v4.jsonl.gz b/corpora/amazon_beauty_compound_association_hair_care.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_compound_association_hair_care_v4.jsonl.gz rename to corpora/amazon_beauty_compound_association_hair_care.jsonl.gz diff --git a/corpora/amazon_beauty_compound_association_makeup_v4.jsonl.gz b/corpora/amazon_beauty_compound_association_makeup.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_compound_association_makeup_v4.jsonl.gz rename to corpora/amazon_beauty_compound_association_makeup.jsonl.gz diff --git a/corpora/amazon_beauty_compound_association_nails_v4.jsonl.gz b/corpora/amazon_beauty_compound_association_nails.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_compound_association_nails_v4.jsonl.gz rename to corpora/amazon_beauty_compound_association_nails.jsonl.gz diff --git a/corpora/amazon_beauty_compound_association_skin_v4.jsonl.gz b/corpora/amazon_beauty_compound_association_skin.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_compound_association_skin_v4.jsonl.gz rename to corpora/amazon_beauty_compound_association_skin.jsonl.gz diff --git a/corpora/amazon_beauty_group_difference_hair_tools_v4.jsonl.gz b/corpora/amazon_beauty_group_difference_hair_tools.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_group_difference_hair_tools_v4.jsonl.gz rename to corpora/amazon_beauty_group_difference_hair_tools.jsonl.gz diff --git a/corpora/amazon_beauty_group_difference_hair_tools_midrating_v5.jsonl.gz b/corpora/amazon_beauty_group_difference_hair_tools_midrating.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_group_difference_hair_tools_midrating_v5.jsonl.gz rename to corpora/amazon_beauty_group_difference_hair_tools_midrating.jsonl.gz diff --git a/corpora/amazon_beauty_group_difference_makeup_v4.jsonl.gz b/corpora/amazon_beauty_group_difference_makeup.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_group_difference_makeup_v4.jsonl.gz rename to corpora/amazon_beauty_group_difference_makeup.jsonl.gz diff --git a/corpora/amazon_beauty_group_difference_nails_v4.jsonl.gz b/corpora/amazon_beauty_group_difference_nails.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_group_difference_nails_v4.jsonl.gz rename to corpora/amazon_beauty_group_difference_nails.jsonl.gz diff --git a/corpora/amazon_beauty_group_difference_skin_v4.jsonl.gz b/corpora/amazon_beauty_group_difference_skin.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_group_difference_skin_v4.jsonl.gz rename to corpora/amazon_beauty_group_difference_skin.jsonl.gz diff --git a/corpora/amazon_beauty_temporal_change_fragrance_v4.jsonl.gz b/corpora/amazon_beauty_temporal_change_fragrance.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_temporal_change_fragrance_v4.jsonl.gz rename to corpora/amazon_beauty_temporal_change_fragrance.jsonl.gz diff --git a/corpora/amazon_beauty_temporal_change_grooming_v4.jsonl.gz b/corpora/amazon_beauty_temporal_change_grooming.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_temporal_change_grooming_v4.jsonl.gz rename to corpora/amazon_beauty_temporal_change_grooming.jsonl.gz diff --git a/corpora/amazon_beauty_temporal_change_hair_tools_v4.jsonl.gz b/corpora/amazon_beauty_temporal_change_hair_tools.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_temporal_change_hair_tools_v4.jsonl.gz rename to corpora/amazon_beauty_temporal_change_hair_tools.jsonl.gz diff --git a/corpora/amazon_beauty_temporal_change_skin_v4.jsonl.gz b/corpora/amazon_beauty_temporal_change_skin.jsonl.gz similarity index 100% rename from corpora/amazon_beauty_temporal_change_skin_v4.jsonl.gz rename to corpora/amazon_beauty_temporal_change_skin.jsonl.gz diff --git a/corpora/app_reviews_compound_association_emulator_v4.jsonl.gz b/corpora/app_reviews_compound_association_emulator.jsonl.gz similarity index 100% rename from corpora/app_reviews_compound_association_emulator_v4.jsonl.gz rename to corpora/app_reviews_compound_association_emulator.jsonl.gz diff --git a/corpora/app_reviews_compound_association_messaging_v4.jsonl.gz b/corpora/app_reviews_compound_association_messaging.jsonl.gz similarity index 100% rename from corpora/app_reviews_compound_association_messaging_v4.jsonl.gz rename to corpora/app_reviews_compound_association_messaging.jsonl.gz diff --git a/corpora/app_reviews_compound_association_news_v4.jsonl.gz b/corpora/app_reviews_compound_association_news.jsonl.gz similarity index 100% rename from corpora/app_reviews_compound_association_news_v4.jsonl.gz rename to corpora/app_reviews_compound_association_news.jsonl.gz diff --git a/corpora/app_reviews_compound_association_services_v4.jsonl.gz b/corpora/app_reviews_compound_association_services.jsonl.gz similarity index 100% rename from corpora/app_reviews_compound_association_services_v4.jsonl.gz rename to corpora/app_reviews_compound_association_services.jsonl.gz diff --git a/corpora/app_reviews_group_difference_display_v4.jsonl.gz b/corpora/app_reviews_group_difference_display.jsonl.gz similarity index 100% rename from corpora/app_reviews_group_difference_display_v4.jsonl.gz rename to corpora/app_reviews_group_difference_display.jsonl.gz diff --git a/corpora/app_reviews_group_difference_emulators_v4.jsonl.gz b/corpora/app_reviews_group_difference_emulators.jsonl.gz similarity index 100% rename from corpora/app_reviews_group_difference_emulators_v4.jsonl.gz rename to corpora/app_reviews_group_difference_emulators.jsonl.gz diff --git a/corpora/app_reviews_group_difference_files_v4.jsonl.gz b/corpora/app_reviews_group_difference_files.jsonl.gz similarity index 100% rename from corpora/app_reviews_group_difference_files_v4.jsonl.gz rename to corpora/app_reviews_group_difference_files.jsonl.gz diff --git a/corpora/app_reviews_group_difference_messaging_v4.jsonl.gz b/corpora/app_reviews_group_difference_messaging.jsonl.gz similarity index 100% rename from corpora/app_reviews_group_difference_messaging_v4.jsonl.gz rename to corpora/app_reviews_group_difference_messaging.jsonl.gz diff --git a/corpora/app_reviews_group_difference_services_v4.jsonl.gz b/corpora/app_reviews_group_difference_services.jsonl.gz similarity index 100% rename from corpora/app_reviews_group_difference_services_v4.jsonl.gz rename to corpora/app_reviews_group_difference_services.jsonl.gz diff --git a/corpora/app_reviews_temporal_change_accessibility_v4.jsonl.gz b/corpora/app_reviews_temporal_change_accessibility.jsonl.gz similarity index 100% rename from corpora/app_reviews_temporal_change_accessibility_v4.jsonl.gz rename to corpora/app_reviews_temporal_change_accessibility.jsonl.gz diff --git a/corpora/app_reviews_temporal_change_game_v4.jsonl.gz b/corpora/app_reviews_temporal_change_game.jsonl.gz similarity index 100% rename from corpora/app_reviews_temporal_change_game_v4.jsonl.gz rename to corpora/app_reviews_temporal_change_game.jsonl.gz diff --git a/corpora/app_reviews_temporal_change_publishing_v4.jsonl.gz b/corpora/app_reviews_temporal_change_publishing.jsonl.gz similarity index 100% rename from corpora/app_reviews_temporal_change_publishing_v4.jsonl.gz rename to corpora/app_reviews_temporal_change_publishing.jsonl.gz diff --git a/corpora/app_reviews_temporal_change_sky_map_v5.jsonl.gz b/corpora/app_reviews_temporal_change_sky_map.jsonl.gz similarity index 100% rename from corpora/app_reviews_temporal_change_sky_map_v5.jsonl.gz rename to corpora/app_reviews_temporal_change_sky_map.jsonl.gz diff --git a/corpora/cfpb_compound_association_equifax_v4.jsonl.gz b/corpora/cfpb_compound_association_equifax.jsonl.gz similarity index 100% rename from corpora/cfpb_compound_association_equifax_v4.jsonl.gz rename to corpora/cfpb_compound_association_equifax.jsonl.gz diff --git a/corpora/cfpb_compound_association_experian_v4.jsonl.gz b/corpora/cfpb_compound_association_experian.jsonl.gz similarity index 100% rename from corpora/cfpb_compound_association_experian_v4.jsonl.gz rename to corpora/cfpb_compound_association_experian.jsonl.gz diff --git a/corpora/cfpb_compound_association_transunion_v4.jsonl.gz b/corpora/cfpb_compound_association_transunion.jsonl.gz similarity index 100% rename from corpora/cfpb_compound_association_transunion_v4.jsonl.gz rename to corpora/cfpb_compound_association_transunion.jsonl.gz diff --git a/corpora/cfpb_group_difference_ca_v4.jsonl.gz b/corpora/cfpb_group_difference_ca.jsonl.gz similarity index 100% rename from corpora/cfpb_group_difference_ca_v4.jsonl.gz rename to corpora/cfpb_group_difference_ca.jsonl.gz diff --git a/corpora/cfpb_group_difference_fl_v4.jsonl.gz b/corpora/cfpb_group_difference_fl.jsonl.gz similarity index 100% rename from corpora/cfpb_group_difference_fl_v4.jsonl.gz rename to corpora/cfpb_group_difference_fl.jsonl.gz diff --git a/corpora/cfpb_group_difference_ga_v4.jsonl.gz b/corpora/cfpb_group_difference_ga.jsonl.gz similarity index 100% rename from corpora/cfpb_group_difference_ga_v4.jsonl.gz rename to corpora/cfpb_group_difference_ga.jsonl.gz diff --git a/corpora/cfpb_group_difference_ny_v4.jsonl.gz b/corpora/cfpb_group_difference_ny.jsonl.gz similarity index 100% rename from corpora/cfpb_group_difference_ny_v4.jsonl.gz rename to corpora/cfpb_group_difference_ny.jsonl.gz diff --git a/corpora/cfpb_group_difference_tx_v4.jsonl.gz b/corpora/cfpb_group_difference_tx.jsonl.gz similarity index 100% rename from corpora/cfpb_group_difference_tx_v4.jsonl.gz rename to corpora/cfpb_group_difference_tx.jsonl.gz diff --git a/corpora/cfpb_temporal_change_equifax_v4.jsonl.gz b/corpora/cfpb_temporal_change_equifax.jsonl.gz similarity index 100% rename from corpora/cfpb_temporal_change_equifax_v4.jsonl.gz rename to corpora/cfpb_temporal_change_equifax.jsonl.gz diff --git a/corpora/cfpb_temporal_change_experian_refreshed_v5.jsonl.gz b/corpora/cfpb_temporal_change_experian_refreshed.jsonl.gz similarity index 100% rename from corpora/cfpb_temporal_change_experian_refreshed_v5.jsonl.gz rename to corpora/cfpb_temporal_change_experian_refreshed.jsonl.gz diff --git a/corpora/cfpb_temporal_change_other_companies_v4.jsonl.gz b/corpora/cfpb_temporal_change_other_companies.jsonl.gz similarity index 100% rename from corpora/cfpb_temporal_change_other_companies_v4.jsonl.gz rename to corpora/cfpb_temporal_change_other_companies.jsonl.gz diff --git a/corpora/cfpb_temporal_change_transunion_v4.jsonl.gz b/corpora/cfpb_temporal_change_transunion.jsonl.gz similarity index 100% rename from corpora/cfpb_temporal_change_transunion_v4.jsonl.gz rename to corpora/cfpb_temporal_change_transunion.jsonl.gz diff --git a/corpora/nhtsa_compound_association_chevrolet_v4.jsonl.gz b/corpora/nhtsa_compound_association_chevrolet.jsonl.gz similarity index 100% rename from corpora/nhtsa_compound_association_chevrolet_v4.jsonl.gz rename to corpora/nhtsa_compound_association_chevrolet.jsonl.gz diff --git a/corpora/nhtsa_compound_association_jeep_v4.jsonl.gz b/corpora/nhtsa_compound_association_jeep.jsonl.gz similarity index 100% rename from corpora/nhtsa_compound_association_jeep_v4.jsonl.gz rename to corpora/nhtsa_compound_association_jeep.jsonl.gz diff --git a/corpora/nhtsa_compound_association_tesla_v4.jsonl.gz b/corpora/nhtsa_compound_association_tesla.jsonl.gz similarity index 100% rename from corpora/nhtsa_compound_association_tesla_v4.jsonl.gz rename to corpora/nhtsa_compound_association_tesla.jsonl.gz diff --git a/corpora/nhtsa_compound_association_toyota_refreshed_v5.jsonl.gz b/corpora/nhtsa_compound_association_toyota_refreshed.jsonl.gz similarity index 100% rename from corpora/nhtsa_compound_association_toyota_refreshed_v5.jsonl.gz rename to corpora/nhtsa_compound_association_toyota_refreshed.jsonl.gz diff --git a/corpora/nhtsa_group_difference_ford_suvs_v4.jsonl.gz b/corpora/nhtsa_group_difference_ford_suvs.jsonl.gz similarity index 100% rename from corpora/nhtsa_group_difference_ford_suvs_v4.jsonl.gz rename to corpora/nhtsa_group_difference_ford_suvs.jsonl.gz diff --git a/corpora/nhtsa_group_difference_honda_cars_v4.jsonl.gz b/corpora/nhtsa_group_difference_honda_cars.jsonl.gz similarity index 100% rename from corpora/nhtsa_group_difference_honda_cars_v4.jsonl.gz rename to corpora/nhtsa_group_difference_honda_cars.jsonl.gz diff --git a/corpora/nhtsa_group_difference_pickups_refreshed_v5.jsonl.gz b/corpora/nhtsa_group_difference_pickups_refreshed.jsonl.gz similarity index 100% rename from corpora/nhtsa_group_difference_pickups_refreshed_v5.jsonl.gz rename to corpora/nhtsa_group_difference_pickups_refreshed.jsonl.gz diff --git a/corpora/nhtsa_group_difference_sedans_v4.jsonl.gz b/corpora/nhtsa_group_difference_sedans.jsonl.gz similarity index 100% rename from corpora/nhtsa_group_difference_sedans_v4.jsonl.gz rename to corpora/nhtsa_group_difference_sedans.jsonl.gz diff --git a/corpora/nhtsa_group_difference_subaru_v4.jsonl.gz b/corpora/nhtsa_group_difference_subaru.jsonl.gz similarity index 100% rename from corpora/nhtsa_group_difference_subaru_v4.jsonl.gz rename to corpora/nhtsa_group_difference_subaru.jsonl.gz diff --git a/corpora/nhtsa_temporal_change_ford_cars_v4.jsonl.gz b/corpora/nhtsa_temporal_change_ford_cars.jsonl.gz similarity index 100% rename from corpora/nhtsa_temporal_change_ford_cars_v4.jsonl.gz rename to corpora/nhtsa_temporal_change_ford_cars.jsonl.gz diff --git a/corpora/nhtsa_temporal_change_honda_suvs_v4.jsonl.gz b/corpora/nhtsa_temporal_change_honda_suvs.jsonl.gz similarity index 100% rename from corpora/nhtsa_temporal_change_honda_suvs_v4.jsonl.gz rename to corpora/nhtsa_temporal_change_honda_suvs.jsonl.gz diff --git a/corpora/nhtsa_temporal_change_nissan_v4.jsonl.gz b/corpora/nhtsa_temporal_change_nissan.jsonl.gz similarity index 100% rename from corpora/nhtsa_temporal_change_nissan_v4.jsonl.gz rename to corpora/nhtsa_temporal_change_nissan.jsonl.gz diff --git a/manifest.json b/manifest.json index 5726814c7c50ed0ebe00a8103b6bfc3de4b8d37a..d61a77516052188d0301b1b33f9b21174eaa1940 100644 --- a/manifest.json +++ b/manifest.json @@ -1,215 +1,215 @@ { - "version": "textinsightbench-v5.1", + "version": "textinsightbench", "files": { "README.md": { - "sha256": "6a1f52b0dff2c1825fcf2f6e25e9315d0d7c6bcb29682d185ff66648577dc3d4", - "bytes": 5952 + "sha256": "7c56388031188b3f44d004ca88469116442561e74a21c3abedbcfd6b302cfdc2", + "bytes": 5957 }, "README.zh-CN.md": { - "sha256": "2c087596c5aebf8eb49925e1463432c8a9f8cabaee94f1b7f08694d6f3a0c3ef", - "bytes": 3389 + "sha256": "d5048bfee44d524477bf9b926fbfa56029d12757844cabb8b9252911f6cdd0ef", + "bytes": 3378 }, "SOURCES.md": { "sha256": "8e43ca1f94c37b760ed2538118c42d74a5e0c56a4350764d7fd595bc5d50dd24", "bytes": 2082 }, - "corpora/amazon_beauty_compound_association_hair_care_v4.jsonl.gz": { + "corpora/amazon_beauty_compound_association_hair_care.jsonl.gz": { "sha256": "e4eddaf50a700ed6094f04ddc99568ab903792ac52c5e08a55e4f6cbf778336d", "bytes": 78929 }, - "corpora/amazon_beauty_compound_association_makeup_v4.jsonl.gz": { + "corpora/amazon_beauty_compound_association_makeup.jsonl.gz": { "sha256": "418d2433feed78d3894cebcb8d857f7befb6bd341713af57a167133922904583", "bytes": 114769 }, - "corpora/amazon_beauty_compound_association_nails_v4.jsonl.gz": { + "corpora/amazon_beauty_compound_association_nails.jsonl.gz": { "sha256": "d63dab5181db0e3247ff20c5b7081a15f55e892e19447efb79ff1c45f84173aa", "bytes": 115088 }, - "corpora/amazon_beauty_compound_association_skin_v4.jsonl.gz": { + "corpora/amazon_beauty_compound_association_skin.jsonl.gz": { "sha256": "301b4dd5f7e96f47abb8c03f2cb70147ed12c4a975fb62fa8423746f286dd1b4", "bytes": 126838 }, - "corpora/amazon_beauty_group_difference_hair_tools_midrating_v5.jsonl.gz": { - "sha256": "7bf3a1c9537c78e0b6cf1beee22908d19fcef3ed815f79857a8df0494adb9910", - "bytes": 71138 - }, - "corpora/amazon_beauty_group_difference_hair_tools_v4.jsonl.gz": { + "corpora/amazon_beauty_group_difference_hair_tools.jsonl.gz": { "sha256": "0603e2eb2807b292830d973eb385dcd0ba95a390af6f5dd0f7415efe73732db3", "bytes": 90323 }, - "corpora/amazon_beauty_group_difference_makeup_v4.jsonl.gz": { + "corpora/amazon_beauty_group_difference_hair_tools_midrating.jsonl.gz": { + "sha256": "7bf3a1c9537c78e0b6cf1beee22908d19fcef3ed815f79857a8df0494adb9910", + "bytes": 71138 + }, + "corpora/amazon_beauty_group_difference_makeup.jsonl.gz": { "sha256": "25e5a34896f27ed3a86e054808c1ac3823d24dc0815f915e37d2c358a94932e5", "bytes": 91559 }, - "corpora/amazon_beauty_group_difference_nails_v4.jsonl.gz": { + "corpora/amazon_beauty_group_difference_nails.jsonl.gz": { "sha256": "04ffe62594b4d0abe15710fd717ebeefac4e45e6f445bde27bfb1059bc8293ae", "bytes": 89939 }, - "corpora/amazon_beauty_group_difference_skin_v4.jsonl.gz": { + "corpora/amazon_beauty_group_difference_skin.jsonl.gz": { "sha256": "e1443c24e6fce8201361c340d320ea4042617f4382ab253a9a8d207c4d45d029", "bytes": 108980 }, - "corpora/amazon_beauty_temporal_change_fragrance_v4.jsonl.gz": { + "corpora/amazon_beauty_temporal_change_fragrance.jsonl.gz": { "sha256": "f0b8ef66e8475ee334f68cef89c225aba9bc27d25a23b68a704e67c42bea8be4", "bytes": 58679 }, - "corpora/amazon_beauty_temporal_change_grooming_v4.jsonl.gz": { + "corpora/amazon_beauty_temporal_change_grooming.jsonl.gz": { "sha256": "fc74044946083aeaa6d18a49b2174dafeb55ac72d264db09698c778db3226b9c", "bytes": 100465 }, - "corpora/amazon_beauty_temporal_change_hair_tools_v4.jsonl.gz": { + "corpora/amazon_beauty_temporal_change_hair_tools.jsonl.gz": { "sha256": "97692f42b1bd3360e0dc566f43bdbd98eb32eff99ad31ac3176bda62c72083b6", "bytes": 89927 }, - "corpora/amazon_beauty_temporal_change_skin_v4.jsonl.gz": { + "corpora/amazon_beauty_temporal_change_skin.jsonl.gz": { "sha256": "a599b62a475936e0a2dd9463750221c26e0d93ebad93007ddb0eee4c44e1c0a9", "bytes": 101050 }, - "corpora/app_reviews_compound_association_emulator_v4.jsonl.gz": { + "corpora/app_reviews_compound_association_emulator.jsonl.gz": { "sha256": "a7655a57e93d61d8e6fcbc74323e6c7cb3551edb44ae1f8120cba7d8663d7d18", "bytes": 21249 }, - "corpora/app_reviews_compound_association_messaging_v4.jsonl.gz": { + "corpora/app_reviews_compound_association_messaging.jsonl.gz": { "sha256": "df146622c29b7f150f10d17b39aa147f047cde596c16344e25f613abbe379dff", "bytes": 42693 }, - "corpora/app_reviews_compound_association_news_v4.jsonl.gz": { + "corpora/app_reviews_compound_association_news.jsonl.gz": { "sha256": "a541586e7409cd922902b8dda3721eeea7d4f8e3afc2ecdf38d1dbbe1b6c4d8b", "bytes": 24278 }, - "corpora/app_reviews_compound_association_services_v4.jsonl.gz": { + "corpora/app_reviews_compound_association_services.jsonl.gz": { "sha256": "88f6ff1feca6108610ad1cc9d8c285e64370b8fad404c84283d032aa393450e5", "bytes": 35278 }, - "corpora/app_reviews_group_difference_display_v4.jsonl.gz": { + "corpora/app_reviews_group_difference_display.jsonl.gz": { "sha256": "bcd975abf7668f97da3fdc890d0b9aa22c354b30f4a72405be977f0f91ca28df", "bytes": 39884 }, - "corpora/app_reviews_group_difference_emulators_v4.jsonl.gz": { + "corpora/app_reviews_group_difference_emulators.jsonl.gz": { "sha256": "3f5c9c619d7f67699ef85e4dc1fa680909c3be82d46029b7b3469f1b31674c63", "bytes": 37271 }, - "corpora/app_reviews_group_difference_files_v4.jsonl.gz": { + "corpora/app_reviews_group_difference_files.jsonl.gz": { "sha256": "3e368b42a24419a7048cd324ac589efdaf9370d85f9e28d69d53e20b6db93ac1", "bytes": 31765 }, - "corpora/app_reviews_group_difference_messaging_v4.jsonl.gz": { + "corpora/app_reviews_group_difference_messaging.jsonl.gz": { "sha256": "9ece304ebe773d0d3f163f83509250cba9edbfad02c77b04b6180afacc055e23", "bytes": 35983 }, - "corpora/app_reviews_group_difference_services_v4.jsonl.gz": { + "corpora/app_reviews_group_difference_services.jsonl.gz": { "sha256": "6eeb2a116c21363507b9350e202b0625655610a374745e7fb8ced83a1c166d42", "bytes": 34910 }, - "corpora/app_reviews_temporal_change_accessibility_v4.jsonl.gz": { + "corpora/app_reviews_temporal_change_accessibility.jsonl.gz": { "sha256": "065dfd4e00d14f32bb8a9b3b42951f1bb5ea3e6228bbb187a430f80e142a78e7", "bytes": 21920 }, - "corpora/app_reviews_temporal_change_game_v4.jsonl.gz": { + "corpora/app_reviews_temporal_change_game.jsonl.gz": { "sha256": "604c99cfd524ae0c5fd54ed560ad9f30bdc14caaec030552ff1b3277777aabc4", "bytes": 25227 }, - "corpora/app_reviews_temporal_change_publishing_v4.jsonl.gz": { + "corpora/app_reviews_temporal_change_publishing.jsonl.gz": { "sha256": "d1c32acf084891b7ec1a4cf3a445da01ebd96ad87585426b955094318511b402", "bytes": 27075 }, - "corpora/app_reviews_temporal_change_sky_map_v5.jsonl.gz": { + "corpora/app_reviews_temporal_change_sky_map.jsonl.gz": { "sha256": "e1ff80a1fbc71daea8ef2c82461e29d6a7d8258948acf236742f665dc81ad528", "bytes": 20358 }, - "corpora/cfpb_compound_association_equifax_v4.jsonl.gz": { + "corpora/cfpb_compound_association_equifax.jsonl.gz": { "sha256": "e69f5827c448b4d06a946b12693498e24396bedb0c51cadde571fbf1332852be", "bytes": 216419 }, - "corpora/cfpb_compound_association_experian_v4.jsonl.gz": { + "corpora/cfpb_compound_association_experian.jsonl.gz": { "sha256": "9794bdd6c496d54cb1e6c39a27d32b761fd97c215a3caca016224058c5545475", "bytes": 203123 }, - "corpora/cfpb_compound_association_transunion_v4.jsonl.gz": { + "corpora/cfpb_compound_association_transunion.jsonl.gz": { "sha256": "3ca730750df35e081f6dc9b10313f1403b735bcfed0b2acb57b7e6d3c774b67b", "bytes": 198260 }, - "corpora/cfpb_group_difference_ca_v4.jsonl.gz": { + "corpora/cfpb_group_difference_ca.jsonl.gz": { "sha256": "a71889fd0d80351cca9cbe3037ff8f78ffd28df901b523934f47052c4ab8cfa7", "bytes": 165546 }, - "corpora/cfpb_group_difference_fl_v4.jsonl.gz": { + "corpora/cfpb_group_difference_fl.jsonl.gz": { "sha256": "956ca973b1c716fd47d47c4c783c6103b009971327c6601ea8089b012dabe1f5", "bytes": 169662 }, - "corpora/cfpb_group_difference_ga_v4.jsonl.gz": { + "corpora/cfpb_group_difference_ga.jsonl.gz": { "sha256": "dd34175d1223d074acd10185663b02a355438119618716f87019185269c170ab", "bytes": 186011 }, - "corpora/cfpb_group_difference_ny_v4.jsonl.gz": { + "corpora/cfpb_group_difference_ny.jsonl.gz": { "sha256": "4c67f818b625c97fef339d9fae2cb2c6da7f3954be67b4c89a21cf712f384932", "bytes": 182663 }, - "corpora/cfpb_group_difference_tx_v4.jsonl.gz": { + "corpora/cfpb_group_difference_tx.jsonl.gz": { "sha256": "342389725fd318329c262304dd9f4f7e2d23784f11a89d5b6e37eb587315d505", "bytes": 159295 }, - "corpora/cfpb_temporal_change_equifax_v4.jsonl.gz": { + "corpora/cfpb_temporal_change_equifax.jsonl.gz": { "sha256": "6f4e061f5ad4d2dd8c4b9f1216945b58a1943b382eadc23310fbb68f7d7748ed", "bytes": 173539 }, - "corpora/cfpb_temporal_change_experian_refreshed_v5.jsonl.gz": { + "corpora/cfpb_temporal_change_experian_refreshed.jsonl.gz": { "sha256": "a34018f6a8138a27f97ba8517d1914678a7b10baae54468349e0f6405d29b5a6", "bytes": 174149 }, - "corpora/cfpb_temporal_change_other_companies_v4.jsonl.gz": { + "corpora/cfpb_temporal_change_other_companies.jsonl.gz": { "sha256": "f95d4cb507b9f9abeb319d89a9c2da0fad9d15a73dfe03566faf21c34c128361", "bytes": 189421 }, - "corpora/cfpb_temporal_change_transunion_v4.jsonl.gz": { + "corpora/cfpb_temporal_change_transunion.jsonl.gz": { "sha256": "15e0b19d239fd5fcac6886022061ee804759170ad51a9314c760a57746cfecaf", "bytes": 173306 }, - "corpora/nhtsa_compound_association_chevrolet_v4.jsonl.gz": { + "corpora/nhtsa_compound_association_chevrolet.jsonl.gz": { "sha256": "4d6cd70af18f54a25e3dc58b7aaf822cffed08ae28710033820bb6f74c6c764f", "bytes": 152502 }, - "corpora/nhtsa_compound_association_jeep_v4.jsonl.gz": { + "corpora/nhtsa_compound_association_jeep.jsonl.gz": { "sha256": "a4006b33adaa8c42755b801b17c9a054daa5c68570164f923beca76761422586", "bytes": 145520 }, - "corpora/nhtsa_compound_association_tesla_v4.jsonl.gz": { + "corpora/nhtsa_compound_association_tesla.jsonl.gz": { "sha256": "9ede3f7ba9663dd77b16ddc38488bd79653facdea1fa212c1cff5adf186a1ebe", "bytes": 168485 }, - "corpora/nhtsa_compound_association_toyota_refreshed_v5.jsonl.gz": { + "corpora/nhtsa_compound_association_toyota_refreshed.jsonl.gz": { "sha256": "77f8ebafd22360c54a0ebc3e7ac4670d18abadb87b3e95693d7a7e7cde14a629", "bytes": 158355 }, - "corpora/nhtsa_group_difference_ford_suvs_v4.jsonl.gz": { + "corpora/nhtsa_group_difference_ford_suvs.jsonl.gz": { "sha256": "f7cccc93c1b8860d9d024a25d85bbb5e6750341277778dae75775116b763e82b", "bytes": 114408 }, - "corpora/nhtsa_group_difference_honda_cars_v4.jsonl.gz": { + "corpora/nhtsa_group_difference_honda_cars.jsonl.gz": { "sha256": "353e7a9d740023b84c30f64d2e698023c1b406300f665991b38c567e0d3bbf72", "bytes": 122005 }, - "corpora/nhtsa_group_difference_pickups_refreshed_v5.jsonl.gz": { + "corpora/nhtsa_group_difference_pickups_refreshed.jsonl.gz": { "sha256": "b6adb92bd8b62d86128c3ec0a4b9ccce05d2fecb552e932d6d7ab19bcddc067e", "bytes": 118616 }, - "corpora/nhtsa_group_difference_sedans_v4.jsonl.gz": { + "corpora/nhtsa_group_difference_sedans.jsonl.gz": { "sha256": "e5d1a5f8804ccccbb3b1d12b379721fd011df906cfa7bf5b33a925057092609b", "bytes": 112510 }, - "corpora/nhtsa_group_difference_subaru_v4.jsonl.gz": { + "corpora/nhtsa_group_difference_subaru.jsonl.gz": { "sha256": "84f69711cba99ed0c2774ee0517ebf1f0bbe01862c99d8df2de2ccb6542d6d92", "bytes": 100315 }, - "corpora/nhtsa_temporal_change_ford_cars_v4.jsonl.gz": { + "corpora/nhtsa_temporal_change_ford_cars.jsonl.gz": { "sha256": "8ddd37685dd177f0c7688342716f6e70db1e83f9454a4e98a90ce1a722667760", "bytes": 109303 }, - "corpora/nhtsa_temporal_change_honda_suvs_v4.jsonl.gz": { + "corpora/nhtsa_temporal_change_honda_suvs.jsonl.gz": { "sha256": "87ed498b8deec9d14d8249442d7bc0fce6199fa122a6a463d8e09191bbe314b3", "bytes": 127007 }, - "corpora/nhtsa_temporal_change_nissan_v4.jsonl.gz": { + "corpora/nhtsa_temporal_change_nissan.jsonl.gz": { "sha256": "be234fb7296cb7e4c148d67184fc2ac2fb5d653e8fc5485a5a6e8fad0a2da02c", "bytes": 116032 }, @@ -1330,16 +1330,16 @@ "bytes": 4798 }, "protocol.json": { - "sha256": "472ffb75d2c738a99d0e005097a7b9748c714b16144515c61f42d346771092ba", - "bytes": 1089 + "sha256": "85c2e42491e9c533ea0808c05f7177a80de4c02f9941db2e0e8e420764e81235", + "bytes": 1081 }, "release.json": { - "sha256": "3bea2f95f0f04a562ba5702c9232b3f8b6cbbdf76a4f2e13cee7eb7a911498f6", - "bytes": 929 + "sha256": "9a6ac85b150c3c450b2c51ff77bff4926d6d64233bad72c56b97682fc4e3c516", + "bytes": 921 }, "tasks.json": { - "sha256": "d24598e7c536cec8b2e8699ebc0e5ad32fc65c99e608911419c9829649851258", - "bytes": 66201 + "sha256": "b623022ca5470f398b6e9d1659a8f0d48fd6e61fc69c8edac333390d1a80e619", + "bytes": 65901 } } } diff --git a/protocol.json b/protocol.json index b11a1981eac56254a72f6c166e8b92d72f86cf7f..f179179e168968f908a6cbd8be087a70021a7ff6 100644 --- a/protocol.json +++ b/protocol.json @@ -1,5 +1,5 @@ { - "version": "textinsightbench-v5.1", + "version": "textinsightbench", "evaluation_unit": "evidence-backed downstream finding", "task_families": { "group_difference": 20, @@ -15,5 +15,5 @@ "sampling": "Finite metadata-selected corpora; estimates describe supplied reports, not population incidence.", "independence": "Tasks share sources and entities. Document separation does not establish statistical independence.", "security": "Corpus content is untrusted input. Use separate execution environments for participants and organizers.", - "scoring_version": "finding-quality-v1" + "scoring_version": "finding-quality" } diff --git a/release.json b/release.json index 59d385a46e4cdb84eff406c1fa2a2c6ec0d833db..440474bb0f7bec3aa48bdb90b084b5916bb2b216 100644 --- a/release.json +++ b/release.json @@ -1,5 +1,5 @@ { - "version": "textinsightbench-v5.1", + "version": "textinsightbench", "tasks": 50, "evaluation_documents": 24504, "learning_documents": 1379468, @@ -35,5 +35,5 @@ "pool_exclusions": {}, "pool_exclusion_scope": "Evaluation documents and reference-development documents, by ID, normalized text and conservative template fingerprint.", "source_pool_documents": 1379468, - "reference_release": "organizer-reference-v2" + "reference_release": "organizer-reference" } diff --git a/tasks.json b/tasks.json index fb91bfea81d4909eb5feabf296dde98113b20026..bd41056bc75e95d009866d72c117c19425a14e75 100644 --- a/tasks.json +++ b/tasks.json @@ -1,6 +1,6 @@ [ { - "task_id": "amazon_beauty_group_difference_hair_tools_midrating_v5", + "task_id": "amazon_beauty_group_difference_hair_tools_midrating", "source": "amazon_beauty", "kind": "group_difference", "scope_name": "hair tools and brushes", @@ -14,7 +14,7 @@ "question": "Within hair tools and brushes, investigate substantive differences in reported experiences between rating_4_5 and rating_3. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 373, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_group_difference_hair_tools_midrating_v5.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_group_difference_hair_tools_midrating.jsonl.gz", "corpus_sha256": "7bf3a1c9537c78e0b6cf1beee22908d19fcef3ed815f79857a8df0494adb9910", "output_contract": "output.schema.json", "selection_scope": { @@ -24,7 +24,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_group_difference_hair_tools_v4", + "task_id": "amazon_beauty_group_difference_hair_tools", "source": "amazon_beauty", "kind": "group_difference", "scope_name": "hair tools and brushes", @@ -38,7 +38,7 @@ "question": "Within hair tools and brushes, investigate substantive differences in reported experiences between rating_4_5 and rating_1_2. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_group_difference_hair_tools_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_group_difference_hair_tools.jsonl.gz", "corpus_sha256": "0603e2eb2807b292830d973eb385dcd0ba95a390af6f5dd0f7415efe73732db3", "output_contract": "output.schema.json", "selection_scope": { @@ -48,7 +48,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_group_difference_nails_v4", + "task_id": "amazon_beauty_group_difference_nails", "source": "amazon_beauty", "kind": "group_difference", "scope_name": "nail-care products", @@ -62,7 +62,7 @@ "question": "Within nail-care products, investigate substantive differences in reported experiences between rating_4_5 and rating_1_2. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_group_difference_nails_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_group_difference_nails.jsonl.gz", "corpus_sha256": "04ffe62594b4d0abe15710fd717ebeefac4e45e6f445bde27bfb1059bc8293ae", "output_contract": "output.schema.json", "selection_scope": { @@ -72,7 +72,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_group_difference_skin_v4", + "task_id": "amazon_beauty_group_difference_skin", "source": "amazon_beauty", "kind": "group_difference", "scope_name": "skin-care and cleansing products", @@ -86,7 +86,7 @@ "question": "Within skin-care and cleansing products, investigate substantive differences in reported experiences between rating_4_5 and rating_1_2. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_group_difference_skin_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_group_difference_skin.jsonl.gz", "corpus_sha256": "e1443c24e6fce8201361c340d320ea4042617f4382ab253a9a8d207c4d45d029", "output_contract": "output.schema.json", "selection_scope": { @@ -96,7 +96,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_group_difference_makeup_v4", + "task_id": "amazon_beauty_group_difference_makeup", "source": "amazon_beauty", "kind": "group_difference", "scope_name": "makeup products", @@ -110,7 +110,7 @@ "question": "Within makeup products, investigate substantive differences in reported experiences between rating_4_5 and rating_1_2. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_group_difference_makeup_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_group_difference_makeup.jsonl.gz", "corpus_sha256": "25e5a34896f27ed3a86e054808c1ac3823d24dc0815f915e37d2c358a94932e5", "output_contract": "output.schema.json", "selection_scope": { @@ -120,7 +120,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_temporal_change_grooming_v4", + "task_id": "amazon_beauty_temporal_change_grooming", "source": "amazon_beauty", "kind": "temporal_change", "scope_name": "shaving and grooming products", @@ -135,7 +135,7 @@ "question": "Within shaving and grooming products, investigate changes in reported experiences before versus on/after 2019-03-04. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_temporal_change_grooming_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_temporal_change_grooming.jsonl.gz", "corpus_sha256": "fc74044946083aeaa6d18a49b2174dafeb55ac72d264db09698c778db3226b9c", "output_contract": "output.schema.json", "selection_scope": { @@ -145,7 +145,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_temporal_change_fragrance_v4", + "task_id": "amazon_beauty_temporal_change_fragrance", "source": "amazon_beauty", "kind": "temporal_change", "scope_name": "fragrances", @@ -160,7 +160,7 @@ "question": "Within fragrances, investigate changes in reported experiences before versus on/after 2020-04-03. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 332, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_temporal_change_fragrance_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_temporal_change_fragrance.jsonl.gz", "corpus_sha256": "f0b8ef66e8475ee334f68cef89c225aba9bc27d25a23b68a704e67c42bea8be4", "output_contract": "output.schema.json", "selection_scope": { @@ -170,7 +170,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_temporal_change_skin_v4", + "task_id": "amazon_beauty_temporal_change_skin", "source": "amazon_beauty", "kind": "temporal_change", "scope_name": "skin-care and cleansing products", @@ -185,7 +185,7 @@ "question": "Within skin-care and cleansing products, investigate changes in reported experiences before versus on/after 2019-01-08. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_temporal_change_skin_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_temporal_change_skin.jsonl.gz", "corpus_sha256": "a599b62a475936e0a2dd9463750221c26e0d93ebad93007ddb0eee4c44e1c0a9", "output_contract": "output.schema.json", "selection_scope": { @@ -195,7 +195,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_temporal_change_hair_tools_v4", + "task_id": "amazon_beauty_temporal_change_hair_tools", "source": "amazon_beauty", "kind": "temporal_change", "scope_name": "hair tools and brushes", @@ -210,7 +210,7 @@ "question": "Within hair tools and brushes, investigate changes in reported experiences before versus on/after 2019-09-07. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_temporal_change_hair_tools_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_temporal_change_hair_tools.jsonl.gz", "corpus_sha256": "97692f42b1bd3360e0dc566f43bdbd98eb32eff99ad31ac3176bda62c72083b6", "output_contract": "output.schema.json", "selection_scope": { @@ -220,7 +220,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_compound_association_hair_care_v4", + "task_id": "amazon_beauty_compound_association_hair_care", "source": "amazon_beauty", "kind": "compound_association", "scope_name": "hair-care products", @@ -228,7 +228,7 @@ "question": "Within hair-care products, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 416, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_compound_association_hair_care_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_compound_association_hair_care.jsonl.gz", "corpus_sha256": "e4eddaf50a700ed6094f04ddc99568ab903792ac52c5e08a55e4f6cbf778336d", "output_contract": "output.schema.json", "selection_scope": { @@ -238,7 +238,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_compound_association_nails_v4", + "task_id": "amazon_beauty_compound_association_nails", "source": "amazon_beauty", "kind": "compound_association", "scope_name": "nail-care products", @@ -246,7 +246,7 @@ "question": "Within nail-care products, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_compound_association_nails_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_compound_association_nails.jsonl.gz", "corpus_sha256": "d63dab5181db0e3247ff20c5b7081a15f55e892e19447efb79ff1c45f84173aa", "output_contract": "output.schema.json", "selection_scope": { @@ -256,7 +256,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_compound_association_makeup_v4", + "task_id": "amazon_beauty_compound_association_makeup", "source": "amazon_beauty", "kind": "compound_association", "scope_name": "makeup products", @@ -264,7 +264,7 @@ "question": "Within makeup products, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_compound_association_makeup_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_compound_association_makeup.jsonl.gz", "corpus_sha256": "418d2433feed78d3894cebcb8d857f7befb6bd341713af57a167133922904583", "output_contract": "output.schema.json", "selection_scope": { @@ -274,7 +274,7 @@ "split": "evaluation" }, { - "task_id": "amazon_beauty_compound_association_skin_v4", + "task_id": "amazon_beauty_compound_association_skin", "source": "amazon_beauty", "kind": "compound_association", "scope_name": "skin-care and cleansing products", @@ -282,7 +282,7 @@ "question": "Within skin-care and cleansing products, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/amazon_beauty_compound_association_skin_v4.jsonl.gz", + "corpus_path": "corpora/amazon_beauty_compound_association_skin.jsonl.gz", "corpus_sha256": "301b4dd5f7e96f47abb8c03f2cb70147ed12c4a975fb62fa8423746f286dd1b4", "output_contract": "output.schema.json", "selection_scope": { @@ -292,7 +292,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_group_difference_services_v4", + "task_id": "app_reviews_group_difference_services", "source": "app_reviews", "kind": "group_difference", "scope_name": "Google Play Services and Google Authenticator", @@ -306,7 +306,7 @@ "question": "Within Google Play Services and Google Authenticator, investigate substantive differences in reported experiences between com.google.android.gms and com.google.android.apps.authenticator2. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/app_reviews_group_difference_services_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_group_difference_services.jsonl.gz", "corpus_sha256": "6eeb2a116c21363507b9350e202b0625655610a374745e7fb8ced83a1c166d42", "output_contract": "output.schema.json", "selection_scope": { @@ -319,7 +319,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_group_difference_messaging_v4", + "task_id": "app_reviews_group_difference_messaging", "source": "app_reviews", "kind": "group_difference", "scope_name": "Telegram and SMS Backup+", @@ -333,7 +333,7 @@ "question": "Within Telegram and SMS Backup+, investigate substantive differences in reported experiences between org.telegram.messenger and com.zegoggles.smssync. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/app_reviews_group_difference_messaging_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_group_difference_messaging.jsonl.gz", "corpus_sha256": "9ece304ebe773d0d3f163f83509250cba9edbfad02c77b04b6180afacc055e23", "output_contract": "output.schema.json", "selection_scope": { @@ -346,7 +346,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_group_difference_emulators_v4", + "task_id": "app_reviews_group_difference_emulators", "source": "app_reviews", "kind": "group_difference", "scope_name": "PPSSPP and Reicast", @@ -360,7 +360,7 @@ "question": "Within PPSSPP and Reicast, investigate substantive differences in reported experiences between org.ppsspp.ppsspp and com.reicast.emulator. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/app_reviews_group_difference_emulators_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_group_difference_emulators.jsonl.gz", "corpus_sha256": "3f5c9c619d7f67699ef85e4dc1fa680909c3be82d46029b7b3469f1b31674c63", "output_contract": "output.schema.json", "selection_scope": { @@ -373,7 +373,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_group_difference_display_v4", + "task_id": "app_reviews_group_difference_display", "source": "app_reviews", "kind": "group_difference", "scope_name": "AcDisplay and Muzei", @@ -387,7 +387,7 @@ "question": "Within AcDisplay and Muzei, investigate substantive differences in reported experiences between com.achep.acdisplay and net.nurik.roman.muzei. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/app_reviews_group_difference_display_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_group_difference_display.jsonl.gz", "corpus_sha256": "bcd975abf7668f97da3fdc890d0b9aa22c354b30f4a72405be977f0f91ca28df", "output_contract": "output.schema.json", "selection_scope": { @@ -400,7 +400,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_group_difference_files_v4", + "task_id": "app_reviews_group_difference_files", "source": "app_reviews", "kind": "group_difference", "scope_name": "FrostWire and DiskUsage", @@ -414,7 +414,7 @@ "question": "Within FrostWire and DiskUsage, investigate substantive differences in reported experiences between com.frostwire.android and com.google.android.diskusage. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/app_reviews_group_difference_files_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_group_difference_files.jsonl.gz", "corpus_sha256": "3e368b42a24419a7048cd324ac589efdaf9370d85f9e28d69d53e20b6db93ac1", "output_contract": "output.schema.json", "selection_scope": { @@ -427,7 +427,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_temporal_change_sky_map_v5", + "task_id": "app_reviews_temporal_change_sky_map", "source": "app_reviews", "kind": "temporal_change", "scope_name": "com.google.android.stardroid", @@ -442,7 +442,7 @@ "question": "Within com.google.android.stardroid, investigate changes in reported experiences before versus on/after 2016-09-16. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 281, "max_findings": 5, - "corpus_path": "corpora/app_reviews_temporal_change_sky_map_v5.jsonl.gz", + "corpus_path": "corpora/app_reviews_temporal_change_sky_map.jsonl.gz", "corpus_sha256": "e1ff80a1fbc71daea8ef2c82461e29d6a7d8258948acf236742f665dc81ad528", "output_contract": "output.schema.json", "selection_scope": { @@ -454,7 +454,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_temporal_change_publishing_v4", + "task_id": "app_reviews_temporal_change_publishing", "source": "app_reviews", "kind": "temporal_change", "scope_name": "org.wordpress.android", @@ -469,7 +469,7 @@ "question": "Within org.wordpress.android, investigate changes in reported experiences before versus on/after 2016-08-14. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 330, "max_findings": 5, - "corpus_path": "corpora/app_reviews_temporal_change_publishing_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_temporal_change_publishing.jsonl.gz", "corpus_sha256": "d1c32acf084891b7ec1a4cf3a445da01ebd96ad87585426b955094318511b402", "output_contract": "output.schema.json", "selection_scope": { @@ -481,7 +481,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_temporal_change_accessibility_v4", + "task_id": "app_reviews_temporal_change_accessibility", "source": "app_reviews", "kind": "temporal_change", "scope_name": "com.google.android.marvin.talkback", @@ -496,7 +496,7 @@ "question": "Within com.google.android.marvin.talkback, investigate changes in reported experiences before versus on/after 2016-12-27. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 312, "max_findings": 5, - "corpus_path": "corpora/app_reviews_temporal_change_accessibility_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_temporal_change_accessibility.jsonl.gz", "corpus_sha256": "065dfd4e00d14f32bb8a9b3b42951f1bb5ea3e6228bbb187a430f80e142a78e7", "output_contract": "output.schema.json", "selection_scope": { @@ -508,7 +508,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_temporal_change_game_v4", + "task_id": "app_reviews_temporal_change_game", "source": "app_reviews", "kind": "temporal_change", "scope_name": "com.watabou.pixeldungeon", @@ -523,7 +523,7 @@ "question": "Within com.watabou.pixeldungeon, investigate changes in reported experiences before versus on/after 2016-04-02. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 301, "max_findings": 5, - "corpus_path": "corpora/app_reviews_temporal_change_game_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_temporal_change_game.jsonl.gz", "corpus_sha256": "604c99cfd524ae0c5fd54ed560ad9f30bdc14caaec030552ff1b3277777aabc4", "output_contract": "output.schema.json", "selection_scope": { @@ -535,7 +535,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_compound_association_services_v4", + "task_id": "app_reviews_compound_association_services", "source": "app_reviews", "kind": "compound_association", "scope_name": "com.google.android.gms", @@ -543,7 +543,7 @@ "question": "Within com.google.android.gms, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/app_reviews_compound_association_services_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_compound_association_services.jsonl.gz", "corpus_sha256": "88f6ff1feca6108610ad1cc9d8c285e64370b8fad404c84283d032aa393450e5", "output_contract": "output.schema.json", "selection_scope": { @@ -555,7 +555,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_compound_association_messaging_v4", + "task_id": "app_reviews_compound_association_messaging", "source": "app_reviews", "kind": "compound_association", "scope_name": "org.telegram.messenger", @@ -563,7 +563,7 @@ "question": "Within org.telegram.messenger, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/app_reviews_compound_association_messaging_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_compound_association_messaging.jsonl.gz", "corpus_sha256": "df146622c29b7f150f10d17b39aa147f047cde596c16344e25f613abbe379dff", "output_contract": "output.schema.json", "selection_scope": { @@ -575,7 +575,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_compound_association_emulator_v4", + "task_id": "app_reviews_compound_association_emulator", "source": "app_reviews", "kind": "compound_association", "scope_name": "com.opendoorstudios.ds4droid", @@ -583,7 +583,7 @@ "question": "Within com.opendoorstudios.ds4droid, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 297, "max_findings": 5, - "corpus_path": "corpora/app_reviews_compound_association_emulator_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_compound_association_emulator.jsonl.gz", "corpus_sha256": "a7655a57e93d61d8e6fcbc74323e6c7cb3551edb44ae1f8120cba7d8663d7d18", "output_contract": "output.schema.json", "selection_scope": { @@ -595,7 +595,7 @@ "split": "evaluation" }, { - "task_id": "app_reviews_compound_association_news_v4", + "task_id": "app_reviews_compound_association_news", "source": "app_reviews", "kind": "compound_association", "scope_name": "org.npr.android.news", @@ -603,7 +603,7 @@ "question": "Within org.npr.android.news, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 263, "max_findings": 5, - "corpus_path": "corpora/app_reviews_compound_association_news_v4.jsonl.gz", + "corpus_path": "corpora/app_reviews_compound_association_news.jsonl.gz", "corpus_sha256": "a541586e7409cd922902b8dda3721eeea7d4f8e3afc2ecdf38d1dbbe1b6c4d8b", "output_contract": "output.schema.json", "selection_scope": { @@ -615,7 +615,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_group_difference_tx_v4", + "task_id": "cfpb_group_difference_tx", "source": "cfpb", "kind": "group_difference", "scope_name": "TX credit-reporting complaints", @@ -629,7 +629,7 @@ "question": "Within TX credit-reporting complaints, investigate substantive differences in reported experiences between Experian Information Solutions Inc. and TRANSUNION INTERMEDIATE HOLDINGS, INC.. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_group_difference_tx_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_group_difference_tx.jsonl.gz", "corpus_sha256": "342389725fd318329c262304dd9f4f7e2d23784f11a89d5b6e37eb587315d505", "output_contract": "output.schema.json", "selection_scope": { @@ -643,7 +643,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_group_difference_fl_v4", + "task_id": "cfpb_group_difference_fl", "source": "cfpb", "kind": "group_difference", "scope_name": "FL credit-reporting complaints", @@ -657,7 +657,7 @@ "question": "Within FL credit-reporting complaints, investigate substantive differences in reported experiences between EQUIFAX, INC. and Experian Information Solutions Inc.. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_group_difference_fl_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_group_difference_fl.jsonl.gz", "corpus_sha256": "956ca973b1c716fd47d47c4c783c6103b009971327c6601ea8089b012dabe1f5", "output_contract": "output.schema.json", "selection_scope": { @@ -671,7 +671,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_group_difference_ca_v4", + "task_id": "cfpb_group_difference_ca", "source": "cfpb", "kind": "group_difference", "scope_name": "CA credit-reporting complaints", @@ -685,7 +685,7 @@ "question": "Within CA credit-reporting complaints, investigate substantive differences in reported experiences between TRANSUNION INTERMEDIATE HOLDINGS, INC. and EQUIFAX, INC.. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_group_difference_ca_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_group_difference_ca.jsonl.gz", "corpus_sha256": "a71889fd0d80351cca9cbe3037ff8f78ffd28df901b523934f47052c4ab8cfa7", "output_contract": "output.schema.json", "selection_scope": { @@ -699,7 +699,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_group_difference_ny_v4", + "task_id": "cfpb_group_difference_ny", "source": "cfpb", "kind": "group_difference", "scope_name": "NY credit-reporting complaints", @@ -713,7 +713,7 @@ "question": "Within NY credit-reporting complaints, investigate substantive differences in reported experiences between Experian Information Solutions Inc. and EQUIFAX, INC.. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_group_difference_ny_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_group_difference_ny.jsonl.gz", "corpus_sha256": "4c67f818b625c97fef339d9fae2cb2c6da7f3954be67b4c89a21cf712f384932", "output_contract": "output.schema.json", "selection_scope": { @@ -727,7 +727,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_group_difference_ga_v4", + "task_id": "cfpb_group_difference_ga", "source": "cfpb", "kind": "group_difference", "scope_name": "GA credit-reporting complaints", @@ -741,7 +741,7 @@ "question": "Within GA credit-reporting complaints, investigate substantive differences in reported experiences between TRANSUNION INTERMEDIATE HOLDINGS, INC. and Experian Information Solutions Inc.. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_group_difference_ga_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_group_difference_ga.jsonl.gz", "corpus_sha256": "dd34175d1223d074acd10185663b02a355438119618716f87019185269c170ab", "output_contract": "output.schema.json", "selection_scope": { @@ -755,7 +755,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_temporal_change_experian_refreshed_v5", + "task_id": "cfpb_temporal_change_experian_refreshed", "source": "cfpb", "kind": "temporal_change", "scope_name": "experian credit-reporting complaints", @@ -770,7 +770,7 @@ "question": "Within experian credit-reporting complaints, investigate changes in reported experiences before versus on/after 2025-01-16. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_temporal_change_experian_refreshed_v5.jsonl.gz", + "corpus_path": "corpora/cfpb_temporal_change_experian_refreshed.jsonl.gz", "corpus_sha256": "a34018f6a8138a27f97ba8517d1914678a7b10baae54468349e0f6405d29b5a6", "output_contract": "output.schema.json", "selection_scope": { @@ -782,7 +782,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_temporal_change_transunion_v4", + "task_id": "cfpb_temporal_change_transunion", "source": "cfpb", "kind": "temporal_change", "scope_name": "transunion credit-reporting complaints", @@ -797,7 +797,7 @@ "question": "Within transunion credit-reporting complaints, investigate changes in reported experiences before versus on/after 2025-01-17. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_temporal_change_transunion_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_temporal_change_transunion.jsonl.gz", "corpus_sha256": "15e0b19d239fd5fcac6886022061ee804759170ad51a9314c760a57746cfecaf", "output_contract": "output.schema.json", "selection_scope": { @@ -809,7 +809,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_temporal_change_equifax_v4", + "task_id": "cfpb_temporal_change_equifax", "source": "cfpb", "kind": "temporal_change", "scope_name": "equifax credit-reporting complaints", @@ -824,7 +824,7 @@ "question": "Within equifax credit-reporting complaints, investigate changes in reported experiences before versus on/after 2025-01-25. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_temporal_change_equifax_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_temporal_change_equifax.jsonl.gz", "corpus_sha256": "6f4e061f5ad4d2dd8c4b9f1216945b58a1943b382eadc23310fbb68f7d7748ed", "output_contract": "output.schema.json", "selection_scope": { @@ -836,7 +836,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_temporal_change_other_companies_v4", + "task_id": "cfpb_temporal_change_other_companies", "source": "cfpb", "kind": "temporal_change", "scope_name": "other companies credit-reporting complaints", @@ -851,7 +851,7 @@ "question": "Within other companies credit-reporting complaints, investigate changes in reported experiences before versus on/after 2025-01-31. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/cfpb_temporal_change_other_companies_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_temporal_change_other_companies.jsonl.gz", "corpus_sha256": "f95d4cb507b9f9abeb319d89a9c2da0fad9d15a73dfe03566faf21c34c128361", "output_contract": "output.schema.json", "selection_scope": { @@ -865,7 +865,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_compound_association_experian_v4", + "task_id": "cfpb_compound_association_experian", "source": "cfpb", "kind": "compound_association", "scope_name": "Experian Information Solutions Inc.", @@ -873,7 +873,7 @@ "question": "Within Experian Information Solutions Inc., discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/cfpb_compound_association_experian_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_compound_association_experian.jsonl.gz", "corpus_sha256": "9794bdd6c496d54cb1e6c39a27d32b761fd97c215a3caca016224058c5545475", "output_contract": "output.schema.json", "selection_scope": { @@ -885,7 +885,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_compound_association_transunion_v4", + "task_id": "cfpb_compound_association_transunion", "source": "cfpb", "kind": "compound_association", "scope_name": "TRANSUNION INTERMEDIATE HOLDINGS, INC.", @@ -893,7 +893,7 @@ "question": "Within TRANSUNION INTERMEDIATE HOLDINGS, INC., discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/cfpb_compound_association_transunion_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_compound_association_transunion.jsonl.gz", "corpus_sha256": "3ca730750df35e081f6dc9b10313f1403b735bcfed0b2acb57b7e6d3c774b67b", "output_contract": "output.schema.json", "selection_scope": { @@ -905,7 +905,7 @@ "split": "evaluation" }, { - "task_id": "cfpb_compound_association_equifax_v4", + "task_id": "cfpb_compound_association_equifax", "source": "cfpb", "kind": "compound_association", "scope_name": "EQUIFAX, INC.", @@ -913,7 +913,7 @@ "question": "Within EQUIFAX, INC., discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/cfpb_compound_association_equifax_v4.jsonl.gz", + "corpus_path": "corpora/cfpb_compound_association_equifax.jsonl.gz", "corpus_sha256": "e69f5827c448b4d06a946b12693498e24396bedb0c51cadde571fbf1332852be", "output_contract": "output.schema.json", "selection_scope": { @@ -925,7 +925,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_group_difference_ford_suvs_v4", + "task_id": "nhtsa_group_difference_ford_suvs", "source": "nhtsa", "kind": "group_difference", "scope_name": "FORD|ESCAPE versus FORD|EXPLORER", @@ -939,7 +939,7 @@ "question": "Within FORD|ESCAPE versus FORD|EXPLORER, investigate substantive differences in reported experiences between FORD|ESCAPE and FORD|EXPLORER. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/nhtsa_group_difference_ford_suvs_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_group_difference_ford_suvs.jsonl.gz", "corpus_sha256": "f7cccc93c1b8860d9d024a25d85bbb5e6750341277778dae75775116b763e82b", "output_contract": "output.schema.json", "selection_scope": { @@ -952,7 +952,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_group_difference_pickups_refreshed_v5", + "task_id": "nhtsa_group_difference_pickups_refreshed", "source": "nhtsa", "kind": "group_difference", "scope_name": "FORD|F-150 versus RAM|1500", @@ -966,7 +966,7 @@ "question": "Within FORD|F-150 versus RAM|1500, investigate substantive differences in reported experiences between FORD|F-150 and RAM|1500. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/nhtsa_group_difference_pickups_refreshed_v5.jsonl.gz", + "corpus_path": "corpora/nhtsa_group_difference_pickups_refreshed.jsonl.gz", "corpus_sha256": "b6adb92bd8b62d86128c3ec0a4b9ccce05d2fecb552e932d6d7ab19bcddc067e", "output_contract": "output.schema.json", "selection_scope": { @@ -979,7 +979,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_group_difference_sedans_v4", + "task_id": "nhtsa_group_difference_sedans", "source": "nhtsa", "kind": "group_difference", "scope_name": "HYUNDAI|SONATA versus KIA|OPTIMA", @@ -993,7 +993,7 @@ "question": "Within HYUNDAI|SONATA versus KIA|OPTIMA, investigate substantive differences in reported experiences between HYUNDAI|SONATA and KIA|OPTIMA. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 453, "max_findings": 5, - "corpus_path": "corpora/nhtsa_group_difference_sedans_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_group_difference_sedans.jsonl.gz", "corpus_sha256": "e5d1a5f8804ccccbb3b1d12b379721fd011df906cfa7bf5b33a925057092609b", "output_contract": "output.schema.json", "selection_scope": { @@ -1006,7 +1006,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_group_difference_honda_cars_v4", + "task_id": "nhtsa_group_difference_honda_cars", "source": "nhtsa", "kind": "group_difference", "scope_name": "HONDA|ACCORD versus HONDA|CIVIC", @@ -1020,7 +1020,7 @@ "question": "Within HONDA|ACCORD versus HONDA|CIVIC, investigate substantive differences in reported experiences between HONDA|ACCORD and HONDA|CIVIC. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/nhtsa_group_difference_honda_cars_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_group_difference_honda_cars.jsonl.gz", "corpus_sha256": "353e7a9d740023b84c30f64d2e698023c1b406300f665991b38c567e0d3bbf72", "output_contract": "output.schema.json", "selection_scope": { @@ -1033,7 +1033,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_group_difference_subaru_v4", + "task_id": "nhtsa_group_difference_subaru", "source": "nhtsa", "kind": "group_difference", "scope_name": "SUBARU|OUTBACK versus SUBARU|FORESTER", @@ -1047,7 +1047,7 @@ "question": "Within SUBARU|OUTBACK versus SUBARU|FORESTER, investigate substantive differences in reported experiences between SUBARU|OUTBACK and SUBARU|FORESTER. Discover which concrete situations distinguish them, quantify the contrast with both denominators, and examine counterexamples and differences in group composition. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 446, "max_findings": 5, - "corpus_path": "corpora/nhtsa_group_difference_subaru_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_group_difference_subaru.jsonl.gz", "corpus_sha256": "84f69711cba99ed0c2774ee0517ebf1f0bbe01862c99d8df2de2ccb6542d6d92", "output_contract": "output.schema.json", "selection_scope": { @@ -1060,7 +1060,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_temporal_change_nissan_v4", + "task_id": "nhtsa_temporal_change_nissan", "source": "nhtsa", "kind": "temporal_change", "scope_name": "nissan vehicle complaints", @@ -1075,7 +1075,7 @@ "question": "Within nissan vehicle complaints, investigate changes in reported experiences before versus on/after 2022-03-10. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/nhtsa_temporal_change_nissan_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_temporal_change_nissan.jsonl.gz", "corpus_sha256": "be234fb7296cb7e4c148d67184fc2ac2fb5d653e8fc5485a5a6e8fad0a2da02c", "output_contract": "output.schema.json", "selection_scope": { @@ -1085,7 +1085,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_temporal_change_ford_cars_v4", + "task_id": "nhtsa_temporal_change_ford_cars", "source": "nhtsa", "kind": "temporal_change", "scope_name": "ford cars vehicle complaints", @@ -1100,7 +1100,7 @@ "question": "Within ford cars vehicle complaints, investigate changes in reported experiences before versus on/after 2022-08-22. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/nhtsa_temporal_change_ford_cars_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_temporal_change_ford_cars.jsonl.gz", "corpus_sha256": "8ddd37685dd177f0c7688342716f6e70db1e83f9454a4e98a90ce1a722667760", "output_contract": "output.schema.json", "selection_scope": { @@ -1113,7 +1113,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_temporal_change_honda_suvs_v4", + "task_id": "nhtsa_temporal_change_honda_suvs", "source": "nhtsa", "kind": "temporal_change", "scope_name": "honda suvs vehicle complaints", @@ -1128,7 +1128,7 @@ "question": "Within honda suvs vehicle complaints, investigate changes in reported experiences before versus on/after 2022-12-10. Identify a specific changing pattern, quantify both period denominators, and examine whether entity, product, rating or reporting composition could explain it. Dates are reporting/review dates, not necessarily event or software-release dates. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 500, "max_findings": 5, - "corpus_path": "corpora/nhtsa_temporal_change_honda_suvs_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_temporal_change_honda_suvs.jsonl.gz", "corpus_sha256": "87ed498b8deec9d14d8249442d7bc0fce6199fa122a6a463d8e09191bbe314b3", "output_contract": "output.schema.json", "selection_scope": { @@ -1141,7 +1141,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_compound_association_chevrolet_v4", + "task_id": "nhtsa_compound_association_chevrolet", "source": "nhtsa", "kind": "compound_association", "scope_name": "CHEVROLET vehicle complaints", @@ -1149,7 +1149,7 @@ "question": "Within CHEVROLET vehicle complaints, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/nhtsa_compound_association_chevrolet_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_compound_association_chevrolet.jsonl.gz", "corpus_sha256": "4d6cd70af18f54a25e3dc58b7aaf822cffed08ae28710033820bb6f74c6c764f", "output_contract": "output.schema.json", "selection_scope": { @@ -1159,7 +1159,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_compound_association_jeep_v4", + "task_id": "nhtsa_compound_association_jeep", "source": "nhtsa", "kind": "compound_association", "scope_name": "JEEP vehicle complaints", @@ -1167,7 +1167,7 @@ "question": "Within JEEP vehicle complaints, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/nhtsa_compound_association_jeep_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_compound_association_jeep.jsonl.gz", "corpus_sha256": "a4006b33adaa8c42755b801b17c9a054daa5c68570164f923beca76761422586", "output_contract": "output.schema.json", "selection_scope": { @@ -1177,7 +1177,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_compound_association_toyota_refreshed_v5", + "task_id": "nhtsa_compound_association_toyota_refreshed", "source": "nhtsa", "kind": "compound_association", "scope_name": "TOYOTA vehicle complaints", @@ -1185,7 +1185,7 @@ "question": "Within TOYOTA vehicle complaints, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/nhtsa_compound_association_toyota_refreshed_v5.jsonl.gz", + "corpus_path": "corpora/nhtsa_compound_association_toyota_refreshed.jsonl.gz", "corpus_sha256": "77f8ebafd22360c54a0ebc3e7ac4670d18abadb87b3e95693d7a7e7cde14a629", "output_contract": "output.schema.json", "selection_scope": { @@ -1195,7 +1195,7 @@ "split": "evaluation" }, { - "task_id": "nhtsa_compound_association_tesla_v4", + "task_id": "nhtsa_compound_association_tesla", "source": "nhtsa", "kind": "compound_association", "scope_name": "TESLA vehicle complaints", @@ -1203,7 +1203,7 @@ "question": "Within TESLA vehicle complaints, discover a recurring association between two distinct reported experiences or events. Explain the context, quantify the joint table and conditional difference, inspect cases where only one occurs, and distinguish an informative association from a definition that makes co-occurrence inevitable. Submit at most five evidence-backed findings or a justified abstention. Do not merely list common words, broad categories or generic sentiment. Account explicitly for unknown judgments. Do not infer causality or population incidence. Any analysis method is allowed.", "n_documents": 600, "max_findings": 5, - "corpus_path": "corpora/nhtsa_compound_association_tesla_v4.jsonl.gz", + "corpus_path": "corpora/nhtsa_compound_association_tesla.jsonl.gz", "corpus_sha256": "9ede3f7ba9663dd77b16ddc38488bd79653facdea1fa212c1cff5adf186a1ebe", "output_contract": "output.schema.json", "selection_scope": {