datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
LSAT_Questionsdart-20250811LSAT_Set_1lm-eval-results-hkust-nlp-dart-math-llama3-8b-prop2diff-private
Dataset Card for Evaluation run of hkust-nlp/dart-math-llama3-8b-prop2diff
Dataset automatically created during the evaluation run of model hkust-nlp/dart-math-llama3-8b-prop2diff
The dataset is composed of 62 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 4 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing… See the full description on the dataset page: https://huggingface.co/datasets/nyu-dice-lab/lm-eval-results-hkust-nlp-dart-math-llama3-8b-prop2diff-private.LSATQuestions3wams2026-rhea-dilutionDartdoc
Dartdoc - 한국 금융공시 텍스트 데이터셋
한국 금융감독원 전자공시시스템(DART) OpenAPI를 통해 수집한 한국어 LLM 학습용 데이터셋입니다.
사업보고서, 증권신고서 등 공시 문서에서 고품질 한국어 텍스트를 추출하였습니다.
데이터셋 개요
항목
내용
언어
한국어 (ko)
수집 기간
2020년 ~ 2025년
총 레코드 수
256,548건
총 텍스트
약 4.6억 자
평균 청크 길이
약 1,794자
출처
금융감독원 DART OpenAPI
수집 대상
공시 유형
코드
대상 문서
필터 조건
정기공시
A
사업보고서
반기/분기보고서 제외
발행공시
C
증권신고서
정정신고서·집합투자 제외
추출 섹션
문서 전체가 아닌 품질이 높은 본문 섹션만 추출합니다.
섹션
내용
II
사업의 내용
IV
이사의 경영진단 및… See the full description on the dataset page: https://huggingface.co/datasets/chaannwooff/Dartdoc.English-German
Dataset Description:
Dataset Name: English-German Translation Pairs for Machine Learning.
Note: This dataset mainly focuses on formal conversation.
Dataset Overview:
This dataset is a meticulously curated collection of English-German translation pairs, designed specifically for training machine learning models, particularly those focused on machine translation tasks. With a total of 48,400,531 translation pairs, this dataset offers an extensive and diverse set of examples that can… See the full description on the dataset page: https://huggingface.co/datasets/Darth-Vaderr/English-German.SWE-bench-DartGeometryDARTS-datasetPreferenceHack
PreferenceHack
PreferenceHack is a paired preference benchmark for evaluating reward models on reward-hacking-style behaviors, released with the code for Activation Reward Models for Few-Shot Model Alignment.
Repository: https://github.com/SKYWALKERRAY/activation-reward-models
Splits
helpful: 1,000 paired text examples focused on helpfulness and safety-oriented judgments.
length: 1,000 paired text examples where length or verbosity can be a misleading reward… See the full description on the dataset page: https://huggingface.co/datasets/DarthVaderSenior/PreferenceHack.spaceship-game-leaderboard
Spaceship Game - Leaderboard
This dataset contains leaderboard entries for the Spaceship Game on Reachy Mini.
Stats
Entries: 2
Top Score: 600 by Pilot
Last Updated: 2026-08-25
Published by: DarthBinks
Format
The leaderboard.json file contains an array of entries:
Field
Type
Description
score
int
Final game score
name
string
Player name
date
string
ISO 8601 timestamp
waves_completed
int?
Number of waves completed… See the full description on the dataset page: https://huggingface.co/datasets/DarthBinks/spaceship-game-leaderboard.adaption-preference-trace-decisions
PreferenceTrace — Source Corpus and Adaption Export
PreferenceTrace tests exact decision-making under competing preferences, evidence, approvals, abstention requirements, temporal/contextual precedence, and machine-readable citation contracts.
Two explicit lineage artifacts
File
Rows
Role
SHA-256
preferencetrace-source-96.jsonl
96
Canonical PreferenceTrace source corpus
7a447f9bf47c3ea455ed96ec36860360aa0e7b9e2dc604450e3a1c665b52363e… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-preference-trace-decisions.vlm-knowledge-conflicttraining_datasetadaption-goose-governance-broad-seed-v1-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-goose_governance_broad_seed_v1 (augmented)
An English instruction-tuning dataset covering core aspects of governance, including political systems, public policy, international relations, and security. Prompts span multiple task types such as conceptual inquiries, comparative institutional analyses, policy trade-off assessments, and evidence synthesis. Completions provide neutral… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-goose-governance-broad-seed-v1-augmented.dart-halspans
DART Hallucination Spans Dataset
A synthetic hallucination detection dataset derived from DART (Data-Record to Text) structured data. Contains 2,000 samples with LLM-generated responses and span-level hallucination annotations.
Dataset Description
This dataset was created to augment RAGTruth for Data2txt (structured data to text) task coverage. An LLM generates both faithful and intentionally hallucinated responses from DART's structured data triples, then annotates the… See the full description on the dataset page: https://huggingface.co/datasets/llm-semantic-router/dart-halspans.dart-20250811Algebra2dart_llm_tasks
DART-LLM Tasks Dataset
Description
DART-LLM Tasks is a dataset designed for evaluating language models in robotic task planning and coordination through few-shot learning. It contains 102 natural language instructions paired with their corresponding structured task decompositions and execution plans.
Dataset Structure
Total examples: 102
Complexity levels:
L1 (Basic): 47 examples
L2 (Medium): 33 examples
L3 (Complex): 22 examples
Features… See the full description on the dataset page: https://huggingface.co/datasets/YongdongWang/dart_llm_tasks.RCW_2025_Positive_Query_Pairs
The Washington law Benchmark (WLB)
Dataset Summary
The Washington Law Benchmark (WLB) is a large-scale, synthetic dataset designed specifically to advance Legal Information Retrieval (IR) and Semantic Search. It bridges the critical "semantic gap" between natural language (how citizens, local governments, and plain-English users describe legal scenarios) and formal statutory legalese (how laws are actually written).
The dataset contains hundreds of thousands of… See the full description on the dataset page: https://huggingface.co/datasets/Darther/RCW_2025_Positive_Query_Pairs.adaption-materials-science-qa-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-materials_science_qa (augmented)
This dataset comprises instruction and response pairs focused on fundamental and advanced topics in materials science. Content includes questions on material properties, synthesis methods, characterization techniques, and practical engineering applications. Each entry is formatted as a prompt and completion pair designed for training technical… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-materials-science-qa-augmented.adaption-cybergoose-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-CyberGoose (augmented)
This dataset contains instruction-response pairs designed to evaluate deterministic reasoning in cybersecurity governance and policy compliance. Each prompt presents a self-contained fictional policy, an operational scenario, and a single proposed action requiring context-based rule interpretation. Completions consist strictly of a binary label classifying the… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-cybergoose-augmented.CollAdmadaption-lord-vader-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-Lord Vader (augmented)
This dataset consists of instruction-response pairs focused on Python code diagnosis and bug repair. Each prompt provides a Python code snippet containing defects alongside context such as tracebacks, failing tests, or expected behavioral specifications. Responses offer concise diagnoses and executable code fixes that resolve the issues while maintaining… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-lord-vader-augmented.adaption-minimal-diff-proofreading-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-minimal_diff_proofreading (augmented)
This dataset consists of instruction-response pairs designed for minimal-difference proofreading across multiple English text formats, including business reports, technical documents, and casual messages. Prompts present text containing objective spelling, punctuation, grammar, and syntax errors, alongside error-free passages. Responses provide… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-minimal-diff-proofreading-augmented.adaption-better-call-saul-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-Better Call Saul (augmented)
This dataset consists of instruction-response pairs designed for evaluating contract law analysis and text-grounded legal reasoning. Prompts feature self-contained contractual excerpts from commercial, employment, licensing, and lease agreements alongside analytical tasks like clause interpretation, conflict detection, and minimal text repair.… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-better-call-saul-augmented.adaption-precision-agriculture-qa-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-precision_agriculture_qa (augmented)
This dataset consists of instruction and response pairs focused on precision agriculture technologies and methodologies. Queries cover topics such as soil sensor integration, variable rate application, yield mapping, and data-driven farm management. Each response provides detailed technical explanations and operational guidance for optimizing… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-precision-agriculture-qa-augmented.adaption-canuckese-augmented
This dataset is a remastered version prepared using Adaption's Adaptive Data platform.
adaption-Canuckese (augmented)
This dataset comprises English instruction-response pairs curated as source material for Canadian English localization workflows. Prompts cover everyday scenarios involving currency, measurements, spelling, institutions, geography, and regional terminology. The completions provide standard, natural English responses that leave clear opportunities for downstream… See the full description on the dataset page: https://huggingface.co/datasets/darthludious/adaption-canuckese-augmented.