datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ai-agent-security-incidents
AI Agent Security Incident Database v0.1
A structured, machine-readable database of 1405 confirmed AI agent security incidents, collected and classified automatically.
What is this?
Every time an AI agent causes unintended harm — escaping a sandbox, exploiting an API, taking unauthorized actions, exfiltrating data — this database captures it.
This is not a list of theoretical risks. Every entry describes something that actually happened, with a verifiable source… See the full description on the dataset page: https://huggingface.co/datasets/gemmozero/ai-agent-security-incidents.circuitlens-gemma-2-2btranscoder-descriptions-and-evaluations
CircuitLens & WeightLens: Transcoder Descriptions and Evaluations
This dataset contains automatically generated descriptions and evaluation metrics for Gemma-2-2B transcoders, produced using CircuitLens and WeightLens methods.
Methods
CircuitLens: https://github.com/egolimblevskaia/CircuitLens
WeightLens: https://github.com/egolimblevskaia/WeightLens
Dataset Structure
The dataset is organized by layers (0, 4, 7, 10, 12, 15, 18, 21, 23, 25), with each layer… See the full description on the dataset page: https://huggingface.co/datasets/egolimblevskaia/circuitlens-gemma-2-2btranscoder-descriptions-and-evaluations.diffing-stats-gemma-2-2b-crosscoder-l13-mu4.1e-02-lr1e-04 Contains maximum activating examples for all the features of our crosscoder trained on gemma 2 2B layer 13 available here: https://huggingface.co/Butanium/gemma-2-2b-crosscoder-l13-mu4.1e-02-lr1e-04/blob/main/README.md
base_examples.pt contains all the maximum examples of the feature on a subset of validation test of fineweb
chat_examples.pt is the same but for lmsys chat data
chat_base_examples.pt is a merge of the two above files.
All files are of the type dict[int, list[tuple[float… See the full description on the dataset page: https://huggingface.co/datasets/science-of-finetuning/diffing-stats-gemma-2-2b-crosscoder-l13-mu4.1e-02-lr1e-04.CACAPO_E2EThe full dataset card is visible in the JSON file named "original_cacapo_for_e2e_models-02_13_2023_19_30_07", which has been made with GEMs second datacard creation GUI.
AIDE-Chip-15K-gem5-Sims
AIDE-Chip 15K gem5 Simulation Dataset
AIDE-Chip-15K-gem5-Sims is a structured dataset of approximately 15,000 validated RISC-V gem5 simulations covering cache hierarchy design-space exploration (DSE) for single-core processors.
The dataset was generated using gem5's Syscall Emulation (SE) mode and six representative workloads, spanning compute-bound, memory-bound, and irregular access patterns. Each sample maps cache configuration parameters to IPC and L2 miss rate, enabling… See the full description on the dataset page: https://huggingface.co/datasets/uralstech/AIDE-Chip-15K-gem5-Sims.med-gemini-medqa-relabeled
Med-Gemini MedQA Relabelling and Analysis
This repository contains data and code corresponding to the MedQA relabelling
performed as part of [1], specifically for the results in Figure 4b and appendix
C.2.
[1] Khaled Saab, Tao Tu, Wei-Hung Weng, Ryutaro Tanno, David Stutz, Ellery Wulczyn,
Fan Zhang, Tim Strother, Chunjong Park, Elahe Vedadi, Juanma Zambrano Chaves,
Szu-Yeu Hu, Mike Schaekermann, Aishwarya Kamath, Yong Cheng, David G.T. Barrett,
Cathy Cheung, Basil… See the full description on the dataset page: https://huggingface.co/datasets/katielink/med-gemini-medqa-relabeled.gemini-finetune-datasetmultilingual-llm-jokes-4o-claude-gemini
Rapidata Generated Joke Preference Dataset
We collected 1'000'000+ human opinions on the jokes generated by state-of-the-art LLMs to decide which model is the funniest. The labelers are shown a joke in their language and asked to answer 'Yes' or 'No' to the question 'Is this joke funny?'.
It took us less than 5 days to get all of the responses.
The jokes are evenly distributed across 5 languages: English, Arabic, Japanese, Vietnamese, Portuguese and across 4 model… See the full description on the dataset page: https://huggingface.co/datasets/Rapidata/multilingual-llm-jokes-4o-claude-gemini.gemma3n-slicing-configsThis repository contains configurations to slice Gemma 3n E4B, which is enabled thanks to it being a MatFormer.
The E4B model can be sliced into small models, trading off quality and latency/compute requirements.
We recommend exploring the [MatFormer Lab](TODO: add link) to getting started with slicing Gemma 3n E4B yourself.
For each configuration, we calculate the MMLU accuracy.
Although these are not the only configurations possible, they are optimal configurations
identified by calculating… See the full description on the dataset page: https://huggingface.co/datasets/google/gemma3n-slicing-configs.gemma-llm-prompt-recoverySelective-Context-Gemma3-12B-resultsgem-analysis-recgym
gem-analysis-recgym
10 gym exercises dataset with 615 abstracted features
数据集信息
来源路径: datasets/recgym
数据大小: 453.0 MB
用途: 健身动作识别模型训练
使用方法
from huggingface_hub import snapshot_download
# 下载数据集
snapshot_download(
repo_id="yonful/gem-analysis-recgym",
repo_type="dataset",
local_dir="./datasets/recgym"
)
或使用项目中的下载脚本:
python scripts/prepare_data.py --dataset recgym
许可证
请参考原始数据源的许可证要求。
diffing-stats-gemma-2-9b-it-L20-k100-lr1e-04-Crosscodergemini_testgemma_2b_outputs
Gemma 2B Green LLM Experiment Outputs
This dataset repository contains experiment artifacts for Gemma 2B green-LLM runs, including LoRA adapter checkpoints, metrics, predictions, carbon logs, and figures.
Contents
checkpoints/: LoRA adapter checkpoints for CE baseline and joint-loss variants.
metrics/: training histories, SQuAD and MMLU summaries, prediction CSVs, calibration tables, and surrogate weights.
logs/: run histories and carbon summary JSON files.
carbon/:… See the full description on the dataset page: https://huggingface.co/datasets/PhotonTJ/gemma_2b_outputs.Gemini_3.1_202_Task_AI_Exposure_Scores
Gemini 3.1 2026 Task AI Exposure Scores
Dataset Summary
This dataset contains task-level AI exposure labels for O*NET task statements. Each task is classified into one of four categories, E0, E1, E2, or E3, using an updated 2026 Agentic AI Exposure Rubric and a Gemini 3.1 Pro classification pipeline. The labels are designed to capture whether a task can be accelerated by a frontier agentic AI system directly, whether it would require deeper software integration, or… See the full description on the dataset page: https://huggingface.co/datasets/MIT-WAL/Gemini_3.1_202_Task_AI_Exposure_Scores.travel-multi-turn-chat-geminiGEMAabsadiffing-stats-gemma-2-9b-it-DPO-L20-k100-lr1e-04-dpo-simpo-Crosscoder10k_reports_gemma_v2
Dataset Card for Financial Document Analysis Dataset
Dataset Description
This dataset comprises structured conversational entries designed to facilitate the training and evaluation of models that analyze and summarize financial documents. Each entry includes a conversation ID, a specific step in the conversation, a system-generated prompt, a user question, and the corresponding model-generated response.
Fields Overview
conv_id: Unique identifier for each… See the full description on the dataset page: https://huggingface.co/datasets/yatharth97/10k_reports_gemma_v2.gemba_annotation
GEMBA Annotation Dataset
Description
This dataset contains human and machine-generated Multidimensional Quality Metrics (MQM) annotations for machine-translated Thai text. It is intended for evaluating and comparing MT system outputs and error annotation quality across different automated models. The dataset features annotations from three large language models (LLMs): Claude 3.7 Sonnet, Gemini 2.0 Flash, and 4o Mini, alongside human annotations, providing a comprehensive… See the full description on the dataset page: https://huggingface.co/datasets/wasanx/gemba_annotation.Gemini3_India_legal_Benchmark
📘 Indian Law Benchmark — 107 Questions (Strict LLM Evaluation)
📝 Dataset Summary
This dataset contains 107 Indian law questions designed to evaluate factual recall, statutory grounding, and calibrated confidence in large language models.The questions span:
Constitution of India
Indian Penal Code (IPC)
Bharatiya Nyaya Sanhita (BNS, 2023)
Code of Criminal Procedure (CrPC)
Bharatiya Nagarik Suraksha Sanhita (BNSS, 2023)
Indian Evidence Act (IEA)
Bharatiya… See the full description on the dataset page: https://huggingface.co/datasets/adhithyakiran/Gemini3_India_legal_Benchmark.gemeldete-stellen-bv41
gemeldete-stellen-bv41
Unofficial Notice
This repository is a private open-source project and is not an official repository of the Federal Statistical Office of Germany (Destatis).
Data Explanation
Official dataset label: Gemeldete Stellen mit Kalender- und Saisonbereinigung nach BV 4.1
Source CSV: https://www.destatis.de/static/de_/opendata/data/gemeldete_stellen_bv41.csv
Source page:… See the full description on the dataset page: https://huggingface.co/datasets/destatis/gemeldete-stellen-bv41.medical-google-chatgpt-gemini-source-overlap
Google and AI Source Overlap Across 12 Medical Niches
An open, reproducible US dataset comparing explicit ChatGPT and Gemini citations with paired Google organic Top 20 results across 12 medical niches and 432 frozen questions.
Full study: https://rotgar.com/medical/resources/google-top-20-chatgpt-gemini-source-overlap
Version DOI: https://doi.org/10.5281/zenodo.21850734
Version: 1.0
Fieldwork: August 7, 2026
Publication date: August 8, 2026
Market and language: United States… See the full description on the dataset page: https://huggingface.co/datasets/RotgarSett/medical-google-chatgpt-gemini-source-overlap.Augmented_CACAPO_for_E2EThe full dataset information can be found in the JSON file named "augmented_cacapo_for_e2e-02_13_2023_22_17_09", which was created with the interactive dataset creator provided by Huggingface.
max-activating-examples-gemma-2-2b-l13-ckissanediffing-stats-SAE-difference_cb-gemma-2-2b-L13-k100-x8-lr1e-04-local-shufflingdiffing-stats-SAE-base-gemma-2-2b-L13-k100-x32-lr1e-04-local-shufflingIMRAD-sections-clf-gemini-augmented
Dataset Card for IMRAD Classification Dataset (100k Rows)
Dataset Name: IMRAD Classification Dataset (100k Rows)
Dataset Description:
This dataset contains approximately 100,000 sentences extracted from scientific research papers and labeled according to their corresponding IMRAD (Introduction, Methods, Results, and Discussion) sections. The data was initially sourced from the unarXive_imrad_clf dataset on Hugging Face and expanded using data augmentation techniques. This dataset is… See the full description on the dataset page: https://huggingface.co/datasets/stormsidali2001/IMRAD-sections-clf-gemini-augmented.diffing-stats-gemma-2-9b-L20-k100-lr1e-04-base-it-Crosscoder