jiosephlee commited on
Commit
70a461b
·
verified ·
1 Parent(s): e96772e

Publish validated auxiliary views dataset v1.0.0

Browse files
ATTRIBUTION.md ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Source attribution and redistribution decisions
2
+
3
+ The table covers all 36 source documents. `Released` refers only to the cleaned full text; probe rows are retained for every source.
4
+
5
+ | Document | Title / issuing court | Creators | Source | Rights status | Released | Modification or omission notice |
6
+ |---|---|---|---|---|---:|---|
7
+ | `arxiv/1_58` | The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits | Shuming Ma, Hongyu Wang, Lingxiao Ma, Lei Wang, Wenhui Wang, Shaohan Huang, Li Dong, Ruiping Wang, Jilong Xue, Furu Wei | [arXiv:2402.17764](https://arxiv.org/abs/2402.17764) | [arXiv non-exclusive distribution license](https://arxiv.org/licenses/nonexclusive-distrib/1.0/) | No | The license grants distribution rights to arXiv, not a general downstream redistribution right. |
8
+ | `arxiv/BOFT` | Parameter-Efficient Orthogonal Finetuning via Butterfly Factorization | Weiyang Liu, Zeju Qiu, Yao Feng, Yuliang Xiu, Yuxuan Xue, Longhui Yu, Haiwen Feng, Zhen Liu, Juyeon Heo, Songyou Peng, Yandong Wen, Michael J. Black, Adrian Weller, Bernhard Schölkopf | [arXiv:2311.06243](https://arxiv.org/abs/2311.06243) | [arXiv non-exclusive distribution license](https://arxiv.org/licenses/nonexclusive-distrib/1.0/) | No | The license grants distribution rights to arXiv, not a general downstream redistribution right. |
9
+ | `arxiv/ByteLatent` | Byte Latent Transformer: Patches Scale Better Than Tokens | Artidoro Pagnoni, Ram Pasunuru, Pedro Rodriguez, John Nguyen, Benjamin Muller, Margaret Li, Chunting Zhou, Lili Yu, Jason Weston, Luke Zettlemoyer, Gargi Ghosh, Mike Lewis, Ari Holtzman, Srinivasan Iyer | [arXiv:2412.09871](https://arxiv.org/abs/2412.09871) | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | Yes | Cleaned from author-submitted TeX; figures, tables, bibliography, and TeX scaffolding may have been removed or normalized. |
10
+ | `arxiv/DPO` | Direct Preference Optimization: Your Language Model is Secretly a Reward Model | Rafael Rafailov, Archit Sharma, Eric Mitchell, Stefano Ermon, Christopher D. Manning, Chelsea Finn | [arXiv:2305.18290](https://arxiv.org/abs/2305.18290) | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | Yes | Cleaned from author-submitted TeX; figures, tables, bibliography, and TeX scaffolding may have been removed or normalized. |
11
+ | `arxiv/FeatLLM` | Large Language Models Can Automatically Engineer Features for Few-Shot Tabular Learning | Sungwon Han, Jinsung Yoon, Sercan O Arik, Tomas Pfister | [arXiv:2404.09491](https://arxiv.org/abs/2404.09491) | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | Yes | Cleaned from author-submitted TeX; figures, tables, bibliography, and TeX scaffolding may have been removed or normalized. |
12
+ | `arxiv/GRPO` | DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models | Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, Y. K. Li, Y. Wu, Daya Guo | [arXiv:2402.03300](https://arxiv.org/abs/2402.03300) | [arXiv non-exclusive distribution license](https://arxiv.org/licenses/nonexclusive-distrib/1.0/) | No | The license grants distribution rights to arXiv, not a general downstream redistribution right. |
13
+ | `arxiv/GSPO` | Group Sequence Policy Optimization | Chujie Zheng, Shixuan Liu, Mingze Li, Xiong-Hui Chen, Bowen Yu, Chang Gao, Kai Dang, Yuqiong Liu, Rui Men, An Yang, Jingren Zhou, Junyang Lin | [arXiv:2507.18071](https://arxiv.org/abs/2507.18071) | [arXiv non-exclusive distribution license](https://arxiv.org/licenses/nonexclusive-distrib/1.0/) | No | The license grants distribution rights to arXiv, not a general downstream redistribution right. |
14
+ | `arxiv/LongRoPE` | LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens | Yiran Ding, Li Lyna Zhang, Chengruidong Zhang, Yuanyuan Xu, Ning Shang, Jiahang Xu, Fan Yang, Mao Yang | [arXiv:2402.13753](https://arxiv.org/abs/2402.13753) | [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/) | No | The cleaned text is modified material, whose redistribution is not permitted by the NoDerivatives term. |
15
+ | `arxiv/OFT` | Controlling Text-to-Image Diffusion by Orthogonal Finetuning | Zeju Qiu, Weiyang Liu, Haiwen Feng, Yuxuan Xue, Yao Feng, Zhen Liu, Dan Zhang, Adrian Weller, Bernhard Schölkopf | [arXiv:2306.07280](https://arxiv.org/abs/2306.07280) | [arXiv non-exclusive distribution license](https://arxiv.org/licenses/nonexclusive-distrib/1.0/) | No | The license grants distribution rights to arXiv, not a general downstream redistribution right. |
16
+ | `arxiv/QLoRA` | QLoRA: Efficient Finetuning of Quantized LLMs | Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, Luke Zettlemoyer | [arXiv:2305.14314](https://arxiv.org/abs/2305.14314) | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | Yes | Cleaned from author-submitted TeX; figures, tables, bibliography, and TeX scaffolding may have been removed or normalized. |
17
+ | `arxiv/fa3` | FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision | Jay Shah, Ganesh Bikshandi, Ying Zhang, Vijay Thakkar, Pradeep Ramani, Tri Dao | [arXiv:2407.08608](https://arxiv.org/abs/2407.08608) | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | Yes | Cleaned from author-submitted TeX; figures, tables, bibliography, and TeX scaffolding may have been removed or normalized. |
18
+ | `arxiv/xLSTM` | xLSTM: Extended Long Short-Term Memory | Maximilian Beck, Korbinian Pöppel, Markus Spanring, Andreas Auer, Oleksandra Prudnikova, Michael Kopp, Günter Klambauer, Johannes Brandstetter, Sepp Hochreiter | [arXiv:2405.04517](https://arxiv.org/abs/2405.04517) | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) | Yes | Cleaned from author-submitted TeX; figures, tables, bibliography, and TeX scaffolding may have been removed or normalized. |
19
+ | `legal/America_First_Legal_Foundation_v_Jamieson_Greer` | America First Legal Foundation v. Jamieson Greer | U.S. Court of Appeals for the District of Columbia Circuit | [CourtListener opinion:11156802](https://www.courtlistener.com/opinion/10690215/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
20
+ | `legal/Apex_Bank_v_Cc_Serve_Corp` | Apex Bank v. Cc Serve Corp. | U.S. Court of Appeals for the Federal Circuit | [CourtListener opinion:11144211](https://www.courtlistener.com/opinion/10677624/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
21
+ | `legal/Bruce_Cohen_v_Consilio_LLC` | Bruce Cohen v. Consilio, LLC | U.S. Court of Appeals for the Eighth Circuit | [CourtListener opinion:11157943](https://www.courtlistener.com/opinion/10691355/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
22
+ | `legal/Finesse_Wireless_LLC_v_Att_Mobility_LLC` | Finesse Wireless LLC v. At&t Mobility LLC | U.S. Court of Appeals for the Federal Circuit | [CourtListener opinion:11143354](https://www.courtlistener.com/opinion/10676767/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
23
+ | `legal/Foad_Farahi_v_FBI` | Foad Farahi v. FBI | U.S. Court of Appeals for the District of Columbia Circuit | [CourtListener opinion:11148298](https://www.courtlistener.com/opinion/10681711/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
24
+ | `legal/Jimenez_v_Bondi` | Jimenez v. Bondi | U.S. Court of Appeals for the Tenth Circuit | [CourtListener opinion:11158572](https://www.courtlistener.com/opinion/10691984/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
25
+ | `legal/Pacito_v_Trump` | Pacito v. Trump | U.S. Court of Appeals for the Ninth Circuit | [CourtListener opinion:11137435](https://www.courtlistener.com/opinion/10670848/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
26
+ | `legal/Santos_v_Kimmel` | Santos v. Kimmel | U.S. Court of Appeals for the Second Circuit | [CourtListener opinion:11157933](https://www.courtlistener.com/opinion/10691345/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
27
+ | `legal/United_States_v_Constantinescu` | United States v. Constantinescu | U.S. Court of Appeals for the Fifth Circuit | [CourtListener opinion:11154541](https://www.courtlistener.com/opinion/10687954/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
28
+ | `legal/United_States_v_Jaison_Coleman` | United States v. Jaison Coleman | U.S. Court of Appeals for the Seventh Circuit | [CourtListener opinion:11158767](https://www.courtlistener.com/opinion/10692179/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
29
+ | `legal/United_States_v_Justin_Cutbank` | United States v. Justin Cutbank | U.S. Court of Appeals for the Eighth Circuit | [CourtListener opinion:11158585](https://www.courtlistener.com/opinion/10691997/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
30
+ | `legal/Williams_v_GoAuto_Insurance` | Williams v. GoAuto Insurance | U.S. Court of Appeals for the Fifth Circuit | [CourtListener opinion:11157029](https://www.courtlistener.com/opinion/10690441/) | [Public domain (U.S. judicial opinion)](https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf) | Yes | Cleaned from CourtListener opinion text; formatting and non-opinion metadata may have been removed or normalized. |
31
+ | `medical/Acute_neurological_Behets_disease_complicated_by_a_syndrome` | Acute neurological Behçet’s disease complicated by a syndrome of inappropriate secretion of antidiuretic hormone | Eiichi Kakehi, Yukako Matsumoto, Shigehisa Sakurai, Kazuhiko Kotani | [PMC12853264](https://pmc.ncbi.nlm.nih.gov/articles/PMC12853264/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
32
+ | `medical/Autoimmune_heparininduced_thrombocytopenia_following_cardiac` | Autoimmune heparin-induced thrombocytopenia following cardiac surgery | Rosanne St Bernard, Ishac Nazy, Theodore Earl Warkentin | [PMC12926959](https://pmc.ncbi.nlm.nih.gov/articles/PMC12926959/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
33
+ | `medical/Central_hypoventilation_and_choking_episodes_revealing_Chiar` | Central hypoventilation and choking episodes revealing Chiari malformation type I | Ali Mohamed Mousa, Mervat Mahmoud, Ghaida Mubarak AlShuraiaan, Zouheir Ibrahim Bitar | [PMC12887155](https://pmc.ncbi.nlm.nih.gov/articles/PMC12887155/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
34
+ | `medical/Dermatomyositis_masquerading_as_angioedema_a_crucial_differe` | Dermatomyositis masquerading as angioedema: a crucial differential not to overlook | Kevin Chan, Gemma Kwan, Hwei Choo Soh, Louise Evans | [PMC12969778](https://pmc.ncbi.nlm.nih.gov/articles/PMC12969778/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
35
+ | `medical/From_dog_bite_to_dialysis_complementmediated_haemolytic_urae` | From dog bite to dialysis: complement-mediated haemolytic uraemic syndrome | Tobias Evason Fredriksen, Burhan-Ud-Din Mian, Szilveszter Dolgos | [PMC12853265](https://pmc.ncbi.nlm.nih.gov/articles/PMC12853265/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
36
+ | `medical/Immunemediated_necrotising_myopathy_following_semaglutide_tr` | Immune-mediated necrotising myopathy following semaglutide treatment: a contributing factor? | Brian Lam, Eleni Tiniakou, Xinhai Robert Zhang, Shervin Assassi | [PMC12969776](https://pmc.ncbi.nlm.nih.gov/articles/PMC12969776/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
37
+ | `medical/Management_of_atypical_complicated_abdominal_masses_in_the_s` | Management of atypical complicated abdominal masses in the second trimester of pregnancy | Rosalba De Nola, Emma Bassi, Cataldo Paletta, Mario Vicino | [PMC12877741](https://pmc.ncbi.nlm.nih.gov/articles/PMC12877741/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
38
+ | `medical/Monoallelic_PARN_mutation_presenting_as_pancytopenia_hepatic` | Monoallelic PARN mutation presenting as pancytopenia, hepatic fibrosis and idiopathic pulmonary fibrosis | Jinal Soni, Karun Saathveeg Sam, Vinus Taneja, Pooja Khosla | [PMC12958418](https://pmc.ncbi.nlm.nih.gov/articles/PMC12958418/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
39
+ | `medical/Multiphasic_anaphylaxis_in_the_emergency_and_intensive_care` | Multiphasic anaphylaxis in the emergency and intensive care setting | Margret Ann Schneider, Hans de Groot, Jan-Willem Geerts | [PMC12958415](https://pmc.ncbi.nlm.nih.gov/articles/PMC12958415/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
40
+ | `medical/Nontype_1_and_nontype_2_diabetes_in_a_young_man_due_to_novel` | Non-type 1 and non-type 2 diabetes in a young man due to novel mutation in HNF1A | Yi-Chen Liu, Sheng-Chiang Su, Feng-Chih Kuo, Yi-Jen Hung, Yen-Lin Chen | [PMC12718537](https://pmc.ncbi.nlm.nih.gov/articles/PMC12718537/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
41
+ | `medical/Pancreatopleural_fistula_in_childhood` | Pancreatopleural fistula in childhood | Hans Vaish, Vipin Vaish, Shalini Suri, Vasudha Tomar | [PMC12728941](https://pmc.ncbi.nlm.nih.gov/articles/PMC12728941/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
42
+ | `medical/TAVinTAVinTAV_after_treated_endocarditis_procedural_strategy` | TAV-in-TAV-in-TAV after treated endocarditis: procedural strategy, imaging planning and 1-year outcome | Evgeny Vladimirovich Merkulov, Goar Kimovna Arutyunyan, Andrey Sergeevich Tereshchenko, Dmitriy V Pevzner, Iiliya Tarasovich Zyuryaev, Mariya Alekseevna Terenicheva, Olga Nickolaevna Ganeeva, Alexey Vladimirovich Knigin | [PMC12983362](https://pmc.ncbi.nlm.nih.gov/articles/PMC12983362/) | [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) | Yes | Cleaned from PubMed Central full text; formatting, references, figures, tables, and metadata may have been removed or normalized. |
43
+
44
+ For released documents, the stated cleaning operation constitutes the change notice supplied with this redistribution.
LICENSE.md ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # License and rights notice
2
+
3
+ This repository contains material under several different rights regimes. No
4
+ single license applies to every field or file.
5
+
6
+ ## Project-authored material
7
+
8
+ To the extent the dataset authors hold copyright or database rights in the
9
+ original probes, annotations, documentation, and selection/arrangement of this
10
+ dataset, those rights are licensed under the **Creative Commons
11
+ Attribution-NonCommercial 4.0 International license (CC BY-NC 4.0)**:
12
+
13
+ https://creativecommons.org/licenses/by-nc/4.0/
14
+
15
+ This grant does not extend to third-party source documents or excerpts.
16
+
17
+ ## Third-party documents
18
+
19
+ Each released document retains the license or public-domain status recorded in
20
+ its row and in ATTRIBUTION.md. In particular, the medical documents are CC
21
+ BY-NC 4.0, the included arXiv documents are CC BY 4.0, and U.S. judicial
22
+ opinions are identified as public-domain government edicts. Required
23
+ attribution and modification notices must be preserved.
24
+
25
+ ## Short source-derived excerpts
26
+
27
+ Fields such as `raw_knowledge_statement` can contain short excerpts or close
28
+ source-derived statements. Where no affirmative redistribution license exists,
29
+ they are included for nonprofit research, evaluation, criticism, and
30
+ reproducibility under an asserted copyright exception/fair-use rationale. They
31
+ are not relicensed under CC BY-NC 4.0, and no permission is granted beyond the
32
+ rights the dataset authors can lawfully grant.
33
+
34
+ Users are responsible for determining whether their use complies with all
35
+ applicable licenses and laws. This notice is not legal advice.
MANIFEST.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "config_rows": {
3
+ "documents": 30,
4
+ "factual_cloze": 6435,
5
+ "factual_mcqa_5shot": 4515,
6
+ "inference_cloze": 430,
7
+ "inference_mcqa_5shot": 322
8
+ },
9
+ "files": {
10
+ "ATTRIBUTION.md": {
11
+ "bytes": 17354,
12
+ "sha256": "9298e0919adcae7d6f6368be93804848a5000332060fe91ed4afdd698f946f23"
13
+ },
14
+ "LICENSE.md": {
15
+ "bytes": 1576,
16
+ "sha256": "060533589b0453a84ee51e567176f1b0686b93b432dd9643977ae3a9af7e840e"
17
+ },
18
+ "README.md": {
19
+ "bytes": 4161,
20
+ "sha256": "2ac947cd54460ee7af2121ee9fe3525499d54db998dbe802ad860b6861d00b77"
21
+ },
22
+ "data/documents/train.parquet": {
23
+ "bytes": 254959,
24
+ "sha256": "03ff57ce7d62dda28eac38208df2dfd5b139bad3ede8df8f8e0075ee5d22caee"
25
+ },
26
+ "data/factual_cloze/test.parquet": {
27
+ "bytes": 1392498,
28
+ "sha256": "04bcc59cc97c238f224e7c253f1174f2b4a612bb0b3d75e3afeac0db75957b03"
29
+ },
30
+ "data/factual_mcqa_5shot/test.parquet": {
31
+ "bytes": 2529208,
32
+ "sha256": "1bf302501ebcd4d2a255bd9a6747f8d20cf6ac004fb4f6c8f50749ec881be981"
33
+ },
34
+ "data/inference_cloze/test.parquet": {
35
+ "bytes": 137067,
36
+ "sha256": "592737abadcb864abbf1d0a988effc615ba8dca1b189086d0f8b8e2211aacfb3"
37
+ },
38
+ "data/inference_mcqa_5shot/test.parquet": {
39
+ "bytes": 312367,
40
+ "sha256": "583332a5bd3cd679c5f795c2eda0261a8f178840c77917342386d0e1a88597bc"
41
+ }
42
+ },
43
+ "release": "v1.0.0",
44
+ "repo_id": "jiosephlee/auxiliary-views-knowledge-acquisition"
45
+ }
README.md ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: other
5
+ license_name: mixed
6
+ license_link: https://huggingface.co/datasets/jiosephlee/auxiliary-views-knowledge-acquisition/blob/main/LICENSE.md
7
+ pretty_name: Auxiliary Views Knowledge Acquisition
8
+ size_categories:
9
+ - 10K<n<100K
10
+ task_categories:
11
+ - text-generation
12
+ - question-answering
13
+ tags:
14
+ - knowledge-probing
15
+ - multiple-choice
16
+ - evaluation
17
+ configs:
18
+ - config_name: documents
19
+ data_files:
20
+ - split: train
21
+ path: data/documents/train.parquet
22
+ - config_name: factual_cloze
23
+ data_files:
24
+ - split: test
25
+ path: data/factual_cloze/test.parquet
26
+ - config_name: factual_mcqa_5shot
27
+ data_files:
28
+ - split: test
29
+ path: data/factual_mcqa_5shot/test.parquet
30
+ - config_name: inference_cloze
31
+ data_files:
32
+ - split: test
33
+ path: data/inference_cloze/test.parquet
34
+ - config_name: inference_mcqa_5shot
35
+ data_files:
36
+ - split: test
37
+ path: data/inference_mcqa_5shot/test.parquet
38
+ ---
39
+
40
+ # Auxiliary Views Knowledge Acquisition
41
+
42
+ This repository contains the cleaned source documents and canonical evaluation
43
+ probes used in *Knowledge Acquisition During Pre-training? Large Language Models
44
+ Learn Better With Auxiliary Views*.
45
+
46
+ ## Configurations
47
+
48
+ | Configuration | Split | Rows | Canonical source |
49
+ |---|---:|---:|---|
50
+ | `documents` | train | 30 | Redistributable cleaned documents |
51
+ | `factual_cloze` | test | 6,435 | factual v14 short-target probes |
52
+ | `factual_mcqa_5shot` | test | 4,515 | factual MCQA v15 |
53
+ | `inference_cloze` | test | 430 | reviewed inference v11 |
54
+ | `inference_mcqa_5shot` | test | 322 | inference MCQA v14 |
55
+
56
+ The factual MCQA configuration is **v15**, the version used for the reported
57
+ evaluation. Paraphrased factual probes and the later v16 MCQA repair are not
58
+ part of this release. Both MCQA configurations contain the exact five-shot
59
+ prompt in `formatted_question_5shot`.
60
+
61
+ Every probe has a stable `probe_id`, a `document_id`, source version, source URL,
62
+ and `document_released`. Factual probes additionally retain the original
63
+ `raw_knowledge_statement`. The latter is a short source-derived statement, not
64
+ the full section or subsection text.
65
+
66
+ ## Copyright and source documents
67
+
68
+ This is a mixed-rights repository; see [LICENSE.md](LICENSE.md) and
69
+ [ATTRIBUTION.md](ATTRIBUTION.md). The 30 released documents comprise:
70
+
71
+ - 6 arXiv papers licensed CC BY 4.0;
72
+ - 12 BMJ Case Reports articles licensed CC BY-NC 4.0; and
73
+ - 12 public-domain U.S. federal judicial opinions obtained via CourtListener.
74
+
75
+ Six cleaned arXiv texts are intentionally omitted:
76
+
77
+ | Local document | arXiv ID | Reason |
78
+ |---|---|---|
79
+ | `1_58` | [2402.17764](https://arxiv.org/abs/2402.17764) | arXiv distribution license does not grant a general downstream redistribution right |
80
+ | `BOFT` | [2311.06243](https://arxiv.org/abs/2311.06243) | same |
81
+ | `GRPO` | [2402.03300](https://arxiv.org/abs/2402.03300) | same |
82
+ | `GSPO` | [2507.18071](https://arxiv.org/abs/2507.18071) | same |
83
+ | `OFT` | [2306.07280](https://arxiv.org/abs/2306.07280) | same |
84
+ | `LongRoPE` | [2402.13753](https://arxiv.org/abs/2402.13753) | CC BY-NC-ND does not permit redistribution of the cleaned/modified text |
85
+
86
+ Probe rows for all 36 documents are retained for exact benchmark reproduction.
87
+ Short third-party excerpts associated with omitted documents are included for
88
+ research and criticism under an asserted copyright exception/fair-use rationale
89
+ and are expressly excluded from the license granted over project-authored work.
90
+ This is a risk-managed research release, not legal advice.
91
+
92
+ ## Intended use and limitations
93
+
94
+ The probes are evaluation data, not training/validation splits. Medical source
95
+ documents are case reports and must not be used for diagnosis, treatment, or
96
+ clinical decision-making. The corpus is small, English-only, intentionally
97
+ heterogeneous, and is not representative of any of its three domains.
98
+
99
+ ## Citation
100
+
101
+ ```bibtex
102
+ @misc{auxiliary_views_knowledge_acquisition,
103
+ title={Knowledge Acquisition During Pre-training? Large Language Models Learn Better With Auxiliary Views},
104
+ author={Dataset Authors},
105
+ year={2026},
106
+ howpublished={Hugging Face dataset},
107
+ url={https://huggingface.co/datasets/jiosephlee/auxiliary-views-knowledge-acquisition}
108
+ }
109
+ ```
data/documents/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03ff57ce7d62dda28eac38208df2dfd5b139bad3ede8df8f8e0075ee5d22caee
3
+ size 254959
data/factual_cloze/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04bcc59cc97c238f224e7c253f1174f2b4a612bb0b3d75e3afeac0db75957b03
3
+ size 1392498
data/factual_mcqa_5shot/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1bf302501ebcd4d2a255bd9a6747f8d20cf6ac004fb4f6c8f50749ec881be981
3
+ size 2529208
data/inference_cloze/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:592737abadcb864abbf1d0a988effc615ba8dca1b189086d0f8b8e2211aacfb3
3
+ size 137067
data/inference_mcqa_5shot/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:583332a5bd3cd679c5f795c2eda0261a8f178840c77917342386d0e1a88597bc
3
+ size 312367