datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
task249_enhanced_wsc_pronoun_disambiguation
Dataset Card for Natural Instructions (https://github.com/allenai/natural-instructions) Task: task249_enhanced_wsc_pronoun_disambiguation
Additional Information
Citation Information
The following paper introduces the corpus in detail. If you use the corpus in published work, please cite it:
@misc{wang2022supernaturalinstructionsgeneralizationdeclarativeinstructions,
title={Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP… See the full description on the dataset page: https://huggingface.co/datasets/Lots-of-LoRAs/task249_enhanced_wsc_pronoun_disambiguation.task275_enhanced_wsc_paraphrase_generation
Dataset Card for Natural Instructions (https://github.com/allenai/natural-instructions) Task: task275_enhanced_wsc_paraphrase_generation
Additional Information
Citation Information
The following paper introduces the corpus in detail. If you use the corpus in published work, please cite it:
@misc{wang2022supernaturalinstructionsgeneralizationdeclarativeinstructions,
title={Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP… See the full description on the dataset page: https://huggingface.co/datasets/Lots-of-LoRAs/task275_enhanced_wsc_paraphrase_generation.python_enhancement_proposals_filtered
Python Enhancement Proposals
Description
Python Enhancement Proposals, or PEPs, are design documents that generally provide a technical specification and rationale for new features of the Python programming language.
There have been 661 PEPs published.
The majority of PEPs are published in the Public Domain, but 5 were published under the “Open Publication License” and omitted from this dataset.
PEPs are long, highly-polished, and technical in nature and often include… See the full description on the dataset page: https://huggingface.co/datasets/common-pile/python_enhancement_proposals_filtered.python_enhancement_proposals
Python Enhancement Proposals
Description
Python Enhancement Proposals, or PEPs, are design documents that generally provide a technical specification and rationale for new features of the Python programming language.
There are been 661 PEPs published.
The majority of PEPs are published in the Public Domain, but 5 were published under the “Open Publication License” and omitted from this dataset.
PEPs are long, highly-polished, and technical in nature and often include… See the full description on the dataset page: https://huggingface.co/datasets/common-pile/python_enhancement_proposals.task276_enhanced_wsc_classification
Dataset Card for Natural Instructions (https://github.com/allenai/natural-instructions) Task: task276_enhanced_wsc_classification
Additional Information
Citation Information
The following paper introduces the corpus in detail. If you use the corpus in published work, please cite it:
@misc{wang2022supernaturalinstructionsgeneralizationdeclarativeinstructions,
title={Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks}… See the full description on the dataset page: https://huggingface.co/datasets/Lots-of-LoRAs/task276_enhanced_wsc_classification.swerebench-traces-raw-source-verification-enhanced-20260617
SWE-rebench Raw Source Verification Enhanced 20260617
This is a private raw source dataset for building refined mini-swe-agent SFT datasets. It is intentionally not tokenized and intentionally preserves source data plus metadata for downstream filtering, masking, weighting, and audit. Do not treat every row as a clean endpoint solve.
Download
The full dataset directory is uploaded as a single compressed archive:
hf download… See the full description on the dataset page: https://huggingface.co/datasets/eewer/swerebench-traces-raw-source-verification-enhanced-20260617.ashaar-with-enhanced-descriptions-baseform-final-sft-lte20-min500-splits
Ashaar Enhanced Description SFT Stratified Splits
Source dataset:
Shaer-AI/ashaar-with-enhanced-descriptions-baseform-final-sft-lte20-min500
Target dataset:
Shaer-AI/ashaar-with-enhanced-descriptions-baseform-final-sft-lte20-min500-splits
This dataset publishes deterministic train / eval / test splits with a 94 / 3 / 3 policy.
Split policy
Primary stratification key:
base_meter
form
length_bucket
Length buckets:
1-3
4-6
7-10
11-20
Small groups fall back… See the full description on the dataset page: https://huggingface.co/datasets/Shaer-AI/ashaar-with-enhanced-descriptions-baseform-final-sft-lte20-min500-splits.swebench-enhanced-cwe
SWE-bench Enhanced with CWE Security Hints
这是 SWE-bench Verified 数据集的增强版本，包含了详细的 CWE（Common Weakness Enumeration）安全提示。
数据集描述
任务: astropy__astropy-12907
仓库: astropy/astropy
Hints 长度: 0 字符
增强内容
原始 SWE-bench 任务的 hints 字段已被增强，包含：
任务特定提示: 指向可能的 bug 位置和修复方向
CWE-754: Improper Check for Unusual or Exceptional Conditions（异常条件检查不足）
CWE-682: Incorrect Calculation（计算错误）
每个 CWE 包含：
详细描述
缓解措施
代码示例
最佳实践
CWE 覆盖
本数据集中的任务映射到以下 CWE：
CWE-754… See the full description on the dataset page: https://huggingface.co/datasets/Chenyang200/swebench-enhanced-cwe.ashaar-with-enhanced-descriptions-baseform-final-sft-lte20-min500
Ashaar Final SFT Dataset with Enhanced Descriptions
This dataset is derived from Shaer-AI/ashaar-with-descriptions-baseform-final-trimmed and is intended to be the final SFT-ready dataset we continue working with.
We got here the hard way. GRPO did not deliver a convincing improvement. Continuation SFT degraded. A fresh-from-zero SFT direction still exposed a deeper data problem. After inspecting the conditioning text, we concluded that many of the old descriptions were weak or… See the full description on the dataset page: https://huggingface.co/datasets/Shaer-AI/ashaar-with-enhanced-descriptions-baseform-final-sft-lte20-min500.ccisd-teks-enhanced
CCISD TEKS Enhanced (LLM-generated)
4,224 records built from the same 428 TEKS expectations as
ccisd-teks-training,
with additional LLM-written fields: detailed explanations, real-world applications,
prerequisite knowledge, common misconceptions, teaching strategies, assessment examples,
cross-curricular connections, and learning progressions.
The added content is LLM output and was not reviewed
The enrichment fields were generated by a language model. No educator… See the full description on the dataset page: https://huggingface.co/datasets/robworks-software/ccisd-teks-enhanced.NuminaMath-Enhanced-CoT-JA-50K
NuminaMath Enhanced CoT Dataset (Japanese 50k Subset)
This repository provides a reasoning-enhanced Japanese math dataset derived from the NuminaMath CoT dataset. The goal is to reinforce the reasoning process in Japanese by prompting a large language model to repeatedly reconsider its steps before arriving at a final answer. This new dataset is not meant to replace the original NuminaMath CoT dataset, but rather to serve as a complementary resource that focuses on multistep… See the full description on the dataset page: https://huggingface.co/datasets/Inoichan/NuminaMath-Enhanced-CoT-JA-50K.math500-enhanced
Math500 Enhanced Dataset
This dataset contains LLM-enhanced versions of mathematical problems with step-by-step reasoning solutions.
Dataset Statistics
Examples: 500 (500 enhanced with LLM)
Enhancement Rate: 100.0%
Data Fields
question: The mathematical problem statement
solution: LLM-enhanced step-by-step solution
original_solution: Original solution text (for reference)
answer: Final numerical answer
level: Problem difficulty level
type: Problem… See the full description on the dataset page: https://huggingface.co/datasets/rachitbansal-harvard/math500-enhanced.deepseek-tui-enhanced-skills
DeepSeek-TUI Enhanced Skills
Structured behavioral skill definitions for DeepSeek-TUI, the terminal-native coding agent for DeepSeek V4.
These skills use a structured ::GENE{} syntax instead of natural language instructions, achieving 35-45% token reduction while reducing interpretation ambiguity.
What's in this dataset
/skills/ — 5 behavioral skill definitions
Skill
What it does
DeepSeek-TUI feature it leverages
session-guardian
Context budget… See the full description on the dataset page: https://huggingface.co/datasets/i-Lang/deepseek-tui-enhanced-skills.tropt-jailbreak-enhancebench-triggers
TROPT — Jailbreak EnhanceBench Triggers (Exp2: enhancement benchmark)
The companion to
tropt-optbench-triggers,
and its mirror image.
sweeps
holds fixed
tropt-optbench-triggers (Exp1)
the optimizer (15 of them)
the recipe: PrefillCE, plain suffix
this dataset (Exp2)
the jailbreak enhancement
the optimizer: always MAC
So Exp1 asks "which search algorithm finds the best trigger?" and Exp2 asks
"given a fixed search algorithm, which jailbreak tricks actually… See the full description on the dataset page: https://huggingface.co/datasets/MatanBT/tropt-jailbreak-enhancebench-triggers.Prompt-Enhancement-Mini