hotchpotch/bekko-system-one-dataset-v0

Bekko System One dataset v0

This dataset is used to train the bekko-system-one-v0 model family.

Public Hugging Face dataset distribution.

Loading and storage

from datasets import load_dataset

dataset = load_dataset("hotchpotch/bekko-system-one-dataset-v0", "laya__typed_decisions")
test = dataset["test"]

Select a configuration explicitly. Parquet uses Zstandard compression and dictionary encoding; all decoded values and row order match the local Arrow release. Only training-manifest.json defines active membership. Configurations listed in quarantine-manifest.json are retained for audit and must not enter evaluation. Source attribution and license scope are retained in sources.json, SOURCES.md, dataset_metadata.json and _sources/. README has one leading YAML block listing every config and split. Hub manifests point to Parquet files; _sources/local-arrow/ preserves the historical Arrow manifests.

Typed decisions over shared state, with Choice, Noul and numeric Score supervision.

Train uses a 300,000-judgment cap per subset with complete cases/groups. Validation is sampled from source validation. Test matches the active S1MB dataset. Use training-manifest.json for membership.

Role	Subsets	Cases	Judgments
train	153	6,589,190	8,421,789
evaluation	125	22,210	41,938
unlabeled	0	0	0

Columns

Column	Purpose
input	Shared state and typed decisions; the only inference input.
targets	Aligned target distributions and annotation metadata.
split	train, validation or test.
case_id	Stable case identifier.
group_id	Related-case grouping.
language	Language code.
input_hash	Stable input identity for audits.

Score values come from the explicit numeric criteria, not candidate indices. Noul retains authored false/true definitions. Preserve soft targets and candidate order. Repeated strings use physical dictionary encoding; decoded values remain strings.

Sources

See sources.json, SOURCES.md and dataset_metadata.json for upstream datasets, revisions, attribution and license scope. The laya__ namespace denotes the Laya-source conversion pipeline; it does not imply Laya authored those upstream datasets. laya__typed_decisions contains four synthetic workflows published by LocalLLaMA. No blanket relicensing is applied. namespace-map.json records names; identifiers remain stable.

Counts below are cases in this release, not upstream dataset sizes or judgment counts. A case can contain multiple judgments. Counts use the active training-manifest.json; 0 means that split is not included. MNLI validation combines validation_matched and validation_mismatched. Source links include acquisition datasets and original dataset references where recorded; they do not imply direct acquisition from each original.

Licenses below describe the original datasets. Follow the links for their full terms. Unknown means the data license has not been confirmed.

Config	Upstream dataset / source	Train	Validation	Test	License / usage terms
abductive_nli	allenai/art	166,937	100	0	Unknown
aegis2	nvidia/Aegis-AI-Content-Safety-Dataset-2.0	27,060	100	100	CC BY 4.0
agnews	sentence-transformers/agnews fancyzhx/ag_news	19,992	0	0	Unknown
all_nli	sentence-transformers/all-nli	20,000	0	0	SNLI: CC BY-SA 4.0; MultiNLI: CC BY 3.0 / CC BY-SA 3.0 / OANC / public domain
aqua_rat	deepmind/aqua_rat	97,113	100	100	Apache-2.0
arc	allenai/ai2_arc	3,352	100	100	CC BY-SA 4.0
arct	tasksource/arct	1,210	100	100	Apache-2.0
argument_quality	ibm-research/argument_quality_ranking_30k	20,972	100	100	CC BY 3.0 (conflicting source notices)
arxiv	hotchpotch/arxiv-ir-simulated-search-queries open-index/open-arxiv	20,000	0	0	CC0 (metadata); paper licenses vary
babi_nli	tasksource/babi_nli	18,590	100	100	BSD
banking77	PolyAI/banking77	10,003	0	100	CC BY 4.0
bbq	lighteval/bbq_helm	0	0	100	CC BY 4.0
bge-hn-colliee	Shitao/bge-m3-data: colliee COLIEE (original benchmark)	463	0	0	Unknown
bge-hn-hotpotqa	Shitao/bge-m3-data: hotpotqa hotpotqa/hotpot_qa	84,176	0	0	CC BY-SA 4.0
bge-hn-miracl_en	Shitao/bge-m3-data: miracl_en miracl/miracl	2,861	0	0	Apache-2.0; Wikipedia text: CC BY-SA / GFDL
bge-hn-mr-tydi_english	Shitao/bge-m3-data: mr-tydi_english castorini/mr-tydi	3,513	0	0	Apache-2.0; Wikipedia text: CC BY-SA / GFDL
bge-hn-msmarco	Shitao/bge-m3-data: msmarco microsoft/ms_marco	300,000	0	0	Non-commercial research only
bge-hn-nli_for_simcse	Shitao/bge-m3-data: nli_for_simcse SimCSE NLI (SNLI + MultiNLI)	274,859	0	0	SNLI: CC BY-SA 4.0; MultiNLI: CC BY 3.0 / CC BY-SA 3.0 / OANC / public domain
bge-hn-nq	Shitao/bge-m3-data: nq google-research-datasets/natural_questions	58,551	0	0	CC BY-SA 3.0
bge-hn-pubmed_qa_labeled	Shitao/bge-m3-data: pubmed_qa_labeled qiaojin/PubMedQA	500	0	0	MIT (dataset); abstract rights vary
bge-hn-squad	Shitao/bge-m3-data: squad rajpurkar/squad	87,526	0	0	CC BY-SA 4.0
bge-hn-trivia	Shitao/bge-m3-data: trivia mandarjoshi/trivia_qa	60,298	0	0	Apache-2.0
bitext	bitext/Bitext-customer-support-llm-chatbot-training-dataset	24,635	0	0	cdla-sharing-1.0
blimp	nyu-mll/blimp	66,999	0	0	CC BY 4.0
boardgameqa	tasksource/Boardgame-QA	15,000	100	100	CC BY 4.0
boolq	google/boolq	7,993	100	0	CC BY-SA 3.0
canttalk	nvidia/CantTalkAboutThis-Topic-Control-Dataset	10,818	0	100	CC BY 4.0
circa	google-research-datasets/circa	34,015	0	0	CC BY 4.0
civil_comments	google/civil_comments	42,857	100	100	CC0-1.0
cladder	https://github.com/causalNLP/cladder	0	0	100	MIT
clinc	clinc/clinc_oos	15,189	151	151	CC BY 3.0
coco_captions	sentence-transformers/coco-captions	20,000	0	0	CC BY 4.0 (captions)
codesearch	nomic-ai/nomic-embed-unsupervised-data CodeSearchNet	19,731	0	0	MIT (tools); source-code licenses vary
commonsense_qa	tau/commonsense_qa	9,741	100	0	MIT
commonsense_qa2	tasksource/commonsense_qa_2.0	9,264	100	0	CC BY 4.0
contract_nli	https://github.com/stanfordnlp/contract-nli	423	61	100	CC BY 4.0
copa	aps/super_glue	399	100	0	BSD-2-Clause (COPA)
corr2cause	tasksource/corr2cause	274,168	100	100	MIT
cosmos_qa	Samsoup/cosmos_qa	25,201	100	0	CC BY 4.0
creak	amydeng2000/CREAK	10,173	100	100	MIT (repository); data license unconfirmed
crows_pairs	https://github.com/nyu-mll/crows-pairs	0	0	100	CC BY-SA 4.0
dbpedia	fancyzhx/dbpedia_14	300,000	0	100	CC BY-SA 3.0
defeasible_nli	tasksource/defeasible-nli	88,448	100	100	Apache-2.0 / CC BY-SA 4.0
esci	tasksource/esci	300,000	0	100	Apache-2.0
ethics	hendrycks/ethics	87,143	100	100	MIT
ethos	SetFit/ethos_binary	598	0	100	GPL-3.0
feasibility_qa	tasksource/feasibilityQA	3,233	0	0	MIT
few_nerd	DFKI-SLT/few-nerd	100,506	100	100	CC BY-SA 4.0
fig_qa	nightingal3/fig-qa	7,962	100	0	MIT
fineweb	hotchpotch/fineweb-ir-simulated-search-queries HuggingFaceFW/fineweb-edu	19,987	0	0	ODC-By 1.0 (database); source-page rights vary
fol_nli	tasksource/FOL-nli	82,188	100	100	Apache-2.0
followir	mteb/Core17InstructionRetrieval mteb/News21InstructionRetrieval mteb/Robust04InstructionRetrieval jhu-clsp/FollowIR-train	1,725	0	0	MIT (annotations); NYT corpus access terms MIT (annotations); Washington Post access terms MIT (annotations); TREC research-access terms Apache-2.0
followir_core17	mteb/Core17InstructionRetrieval	0	0	100	MIT (annotations); NYT corpus access terms
followir_news21	mteb/News21InstructionRetrieval	0	0	100	MIT (annotations); Washington Post access terms
followir_robust04	mteb/Robust04InstructionRetrieval	0	0	100	MIT (annotations); TREC research-access terms
go_emotions	google-research-datasets/go_emotions	10,714	100	100	Apache-2.0
gooaq	sentence-transformers/gooaq allenai/gooaq	19,993	0	0	Non-commercial use only
gretel_pii	gretelai/gretel-pii-masking-en-v1	48,221	100	100	Apache-2.0
gsm8k	openai/gsm8k	7,473	0	100	MIT
hans	tasksource/hans	29,851	100	100	MIT
hatecheck	Paul/hatecheck	0	0	100	CC BY 4.0
helpsteer	nvidia/HelpSteer	32,872	100	0	CC BY 4.0
helpsteer2	nvidia/HelpSteer2	20,321	100	0	CC BY 4.0
helpsteer3	nvidia/HelpSteer3	37,844	100	0	CC BY 4.0
hh_rlhf	Anthropic/hh-rlhf	160,149	0	100	MIT
hwu64	FastFit/hwu_64	7,251	100	100	CC BY 4.0
impli	https://github.com/UKPLab/acl2022-impli	0	0	100	Apache-2.0
laya__ag_news	fancyzhx/ag_news	120,000	0	100	Unknown
laya__enron_spam	SetFit/enron_spam	28,143	0	100	Unknown
laya__phishing	zefang-liu/phishing-email-dataset	17,535	0	0	LGPL-3.0
laya__typed_decisions	LocalLLaMA/typed-decisions	1,200	0	100	Apache-2.0
lexcomp	tasksource/lexcomp-nc-relation	1,272	100	100	Apache-2.0
logical_entailment	tasksource/logical-entailment	99,754	100	100	Apache-2.0
lonli	tasksource/lonli	0	0	100	MIT
massive	AmazonScience/massive	10,463	100	100	CC BY 4.0
measuring_hate	ucberkeley-dlab/measuring-hate-speech	30,000	0	0	CC BY 4.0
minds14	PolyAI/minds14	1,663	0	0	CC BY 4.0
miqa	tasksource/miqa	0	0	100	CC BY 4.0
mnli	nyu-mll/multi_nli	300,000	100	0	CC BY 3.0 / CC BY-SA 3.0 / OANC / public domain
monotonicity	tasksource/monotonicity-entailment	5,379	0	0	Apache-2.0
mtop	tasksource/mtop	15,597	100	100	CC BY-SA 4.0
natural_logic	tasksource/naturallogic	6,386	0	0	Apache-2.0
natural_questions	sentence-transformers/natural-questions google-research-datasets/natural_questions	19,948	0	0	CC BY-SA 3.0
nlsat	tasksource/nlsat	160,000	100	100	Apache-2.0
nlupp	https://github.com/PolyAI-LDN/task-specific-datasets	3,040	0	0	CC BY 4.0
oasst1	OpenAssistant/oasst1	21,843	100	0	Apache-2.0
open_jev__amount-extraction-control-v1	ZefanCai/Open-Jev	32,984	100	100	CC0-1.0
open_jev__browser-control-v1	ZefanCai/Open-Jev	13,814	100	100	CC0-1.0
open_jev__citation-control-v1	ZefanCai/Open-Jev	2,520	100	100	CC0-1.0
open_jev__context-retention-control-v1	ZefanCai/Open-Jev	6,138	100	100	CC0-1.0
open_jev__customer-control-v1	ZefanCai/Open-Jev	4,206	100	100	CC0 (conversations); question license unknown
open_jev__drone-control-v1	ZefanCai/Open-Jev	15,694	100	100	CC0-1.0
open_jev__email-selection-control-v1	ZefanCai/Open-Jev	3,618	100	100	CC0-1.0
open_jev__entity-alignment-control-v1	ZefanCai/Open-Jev	6,944	100	100	CC0-1.0
open_jev__ir-control-v1	ZefanCai/Open-Jev	7,366	100	100	CC0-1.0
open_jev__mailroom-control-v1	ZefanCai/Open-Jev	73,472	100	100	CC0-1.0
open_jev__painting-geometry-v1	ZefanCai/Open-Jev	23,552	100	100	CC0-1.0
open_jev__phone-extraction-control-v1	ZefanCai/Open-Jev	12,350	100	100	CC0-1.0
open_jev__reasoning-control-v1	ZefanCai/Open-Jev	5,442	100	100	CC0-1.0
open_jev__silent-failure-control-v1	ZefanCai/Open-Jev	6,432	100	100	CC0-1.0
open_jev__snake-v1	ZefanCai/Open-Jev	12,508	100	100	CC0-1.0
open_jev__sponsor-segment-control-v1	ZefanCai/Open-Jev	6,345	100	100	CC0-1.0
open_jev__tic_tac_toe-v1	ZefanCai/Open-Jev	3,264	100	100	CC0-1.0
open_jev__tile_platformer-v1	ZefanCai/Open-Jev	1,346	0	0	CC0-1.0
open_jev__trex_runner-v1	ZefanCai/Open-Jev	28	0	0	CC0-1.0
open_jev__vizdoom-basic-v1	ZefanCai/Open-Jev	6,354	100	66	CC0 (numeric data)
open_jev__workflow-controls-v1__agent_trace_observability	ZefanCai/Open-Jev	3,780	100	100	CC0-1.0
open_jev__workflow-controls-v1__customer_service	ZefanCai/Open-Jev	4,392	100	100	CC0-1.0
open_jev__workflow-controls-v1__invoice_processing	ZefanCai/Open-Jev	5,370	100	100	CC0-1.0
open_jev__workflow-controls-v1__security_incidents	ZefanCai/Open-Jev	8,874	100	100	CC0-1.0
openbookqa	allenai/openbookqa	4,870	100	100	Apache-2.0
patent_similarity	tasksource/patent-phrase-similarity	36,473	100	100	CC BY 4.0
paws	google-research-datasets/paws	49,334	100	100	Google custom license; Wikipedia text: CC BY-SA / GFDL
piqa	lighteval/piqa	15,041	100	0	AFL-3.0
plane	tasksource/PLANE-ood	161,073	0	100	CC BY 2.0
poem_sentiment	google-research-datasets/poem_sentiment	892	100	100	CC BY 4.0
pubmed	hotchpotch/pubmed-abstract-ir-simulated-search-queries slinusc/PubMedAbstractsSubset	19,993	0	0	CC0 (metadata); abstract rights vary
qasc	allenai/qasc	8,129	100	0	CC BY 4.0
qasper	allenai/qasper	392	89	100	CC BY 4.0
quartz	allenai/quartz	2,672	100	100	CC BY 4.0
robust_lr	tasksource/robustLR	0	0	100	MIT
ropes	allenai/ropes	6,379	100	0	CC BY 4.0
ruletaker	tasksource/ruletaker	300,000	100	100	Apache-2.0
s1mb-generalization-contextual-choice	Authored evaluation source	0	0	100	not separately declared
s1mb-generalization-contextual-noul	Authored evaluation source	0	0	100	not separately declared
s1mb-generalization-contextual-score	Authored evaluation source	0	0	100	not separately declared
s1mb-generalization-diverse-choice	Authored evaluation source	0	0	100	not separately declared
s1mb-generalization-diverse-noul	Authored evaluation source	0	0	100	not separately declared
s1mb-generalization-diverse-score	Authored evaluation source	0	0	100	not separately declared
scicite	tasksource/scicite	6,099	100	100	Unknown
scitail	allenai/scitail	23,563	100	100	Apache-2.0
scone	tasksource/scone	4,986	0	100	CC0-1.0
sdoh_nli	tasksource/SDOH-NLI	20,465	100	100	CC BY 4.0
sgd	https://github.com/google-research-datasets/dstc8-schema-guided-dialogue	173,481	100	100	CC BY-SA 4.0
sms_spam	ucirvine/sms_spam	5,171	0	0	CC BY 4.0
snips	https://github.com/sonos/nlu-benchmark	13,621	100	0	CC0-1.0
snli	stanfordnlp/snli	300,000	100	100	CC BY-SA 4.0
social_iqa	baber/social_i_qa	33,193	100	0	CC BY 4.0
space_nli	tasksource/SpaceNLI	31,877	0	0	MIT
spartqa	tasksource/spartqa-mchoice	22,472	100	100	MIT
squad2	rajpurkar/squad_v2	129,668	100	0	CC BY-SA 4.0
stepgame	tasksource/stepgame	295,827	100	100	MIT
strategyqa	tasksource/strategy-qa	2,289	0	0	MIT
synthetic_relevance__agnews	sentence-transformers/agnews fancyzhx/ag_news	4,800	0	0	Unknown
synthetic_relevance__all_nli	sentence-transformers/all-nli	4,800	0	0	SNLI: CC BY-SA 4.0; MultiNLI: CC BY 3.0 / CC BY-SA 3.0 / OANC / public domain
synthetic_relevance__arxiv	hotchpotch/arxiv-ir-simulated-search-queries open-index/open-arxiv	4,800	0	0	CC0 (metadata); paper licenses vary
synthetic_relevance__bge_m3__hotpotqa	Shitao/bge-m3-data: hotpotqa hotpotqa/hotpot_qa	4,800	0	0	CC BY-SA 4.0
synthetic_relevance__bge_m3__miracl_en	Shitao/bge-m3-data: miracl_en miracl/miracl	4,800	0	0	Apache-2.0; Wikipedia text: CC BY-SA / GFDL
synthetic_relevance__bge_m3__mr-tydi_english	Shitao/bge-m3-data: mr-tydi_english castorini/mr-tydi	4,800	0	0	Apache-2.0; Wikipedia text: CC BY-SA / GFDL
synthetic_relevance__bge_m3__msmarco	Shitao/bge-m3-data: msmarco microsoft/ms_marco	4,608	0	0	Non-commercial research only
synthetic_relevance__bge_m3__nq	Shitao/bge-m3-data: nq google-research-datasets/natural_questions	4,800	0	0	CC BY-SA 3.0
synthetic_relevance__bge_m3__pubmed_qa_labeled	Shitao/bge-m3-data: pubmed_qa_labeled qiaojin/PubMedQA	4,000	0	0	MIT (dataset); abstract rights vary
synthetic_relevance__bge_m3__squad	Shitao/bge-m3-data: squad rajpurkar/squad	4,800	0	0	CC BY-SA 4.0
synthetic_relevance__bge_m3__trivia	Shitao/bge-m3-data: trivia mandarjoshi/trivia_qa	4,800	0	0	Apache-2.0
synthetic_relevance__coco_captions	sentence-transformers/coco-captions	4,800	0	0	CC BY 4.0 (captions)
synthetic_relevance__codesearch	nomic-ai/nomic-embed-unsupervised-data CodeSearchNet	4,800	0	0	MIT (tools); source-code licenses vary
synthetic_relevance__fineweb	hotchpotch/fineweb-ir-simulated-search-queries HuggingFaceFW/fineweb-edu	4,800	0	0	ODC-By 1.0 (database); source-page rights vary
synthetic_relevance__gooaq	sentence-transformers/gooaq allenai/gooaq	4,800	0	0	Non-commercial use only
synthetic_relevance__nanobeir__NanoArguAna	NanoArguAna	0	0	400	CC BY 4.0
synthetic_relevance__nanobeir__NanoDBPedia	NanoDBPedia	0	0	400	CC BY-SA 3.0 / GFDL
synthetic_relevance__nanobeir__NanoFiQA2018	NanoFiQA2018	0	0	400	Non-commercial use only / CC BY-SA 2.5 / CC BY-SA 3.0
synthetic_relevance__nanobeir__NanoHotpotQA	NanoHotpotQA	0	0	400	CC BY-SA 4.0
synthetic_relevance__nanobeir__NanoMSMARCO	NanoMSMARCO	0	0	400	Non-commercial research only
synthetic_relevance__nanobeir__NanoNFCorpus	NanoNFCorpus	0	0	400	Unknown
synthetic_relevance__nanobeir__NanoNQ	NanoNQ	0	0	400	CC BY-SA 3.0
synthetic_relevance__nanobeir__NanoSCIDOCS	NanoSCIDOCS	0	0	400	CC BY 4.0
synthetic_relevance__nanobeir__NanoTouche2020	NanoTouche2020	0	0	392	CC BY 4.0
synthetic_relevance__nanocoir	NanoApps NanoCodeFeedbackMT NanoCodeFeedbackST NanoCodeSearchNet NanoCodeSearchNetCCR NanoCodeTransOceanContest NanoCodeTransOceanDL NanoCosQA NanoStackOverflowQA NanoSyntheticText2SQL	0	0	800	Non-commercial research only Apache-2.0 Non-commercial research only; source-code licenses vary CC BY-SA 3.0
synthetic_relevance__natural_questions	sentence-transformers/natural-questions google-research-datasets/natural_questions	4,800	0	0	CC BY-SA 3.0
synthetic_relevance__pubmed	hotchpotch/pubmed-abstract-ir-simulated-search-queries slinusc/PubMedAbstractsSubset	4,800	0	0	CC0 (metadata); abstract rights vary
synthetic_relevance__trivia_qa	sentence-transformers/trivia-qa mandarjoshi/trivia_qa	4,800	0	0	Apache-2.0
synthetic_relevance__wikipedia_hard_negatives_english	hotchpotch/wikipedia-english-ir-simulated-search-queries	4,800	0	0	CC BY-SA 4.0 / GFDL
taskmaster	https://github.com/google-research-datasets/Taskmaster	17,295	0	0	CC BY 4.0
temporal_nli	tasksource/temporal-nli	72,322	0	100	Apache-2.0
tracie	tasksource/tracie	1,174	0	100	Apache-2.0
trivia_qa	sentence-transformers/trivia-qa mandarjoshi/trivia_qa	18,775	0	0	Apache-2.0
ud_ewt	universal-dependencies/universal_dependencies	11,663	100	100	CC BY-SA 4.0
wikipedia_hard_negatives_english	hotchpotch/wikipedia-english-ir-simulated-search-queries	300,000	0	0	CC BY-SA 4.0 / GFDL
winobias	https://github.com/uclanlp/corefBias	0	0	100	MIT
winogrande	allenai/winogrande	40,392	100	0	Apache-2.0
winowhy	tasksource/winowhy	0	0	100	MIT
wiqa	allenai/wiqa	29,324	100	100	Apache-2.0

License

Each subset follows its original dataset's license and usage terms. Check the linked original sources before using or redistributing the data. This collection does not change those licenses.

Where noted, code, annotations and source documents have different terms. Detailed source records are in sources.json. An unknown license is not permission to reuse the data.

For Open-Jev, retain its acquisition attribution and third-party notices. In open_jev__customer-control-v1, the generated conversations have a CC0-1.0 declaration, but the upstream question descriptions have no verified license and are not relicensed as CC0. Authored S1MB generalization subsets retain their recorded not separately declared status. See sources.json, SOURCES.md and the retained upstream notices for scope and qualifications.