datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
FACTS-grounding-public
FACTS Grounding 1.0 Public Examples
860 public FACTS Grounding examples from Google DeepMind and Google Research
FACTS Grounding is a benchmark from Google DeepMind and Google Research designed to measure the performance of AI Models on factuality and grounding.
▶ FACTS Grounding Leaderboard on Kaggle▶ Technical Report▶ Evaluation Starter Code▶ Google DeepMind Blog Post
Usage
The FACTS Grounding benchmark evaluates the ability of Large Language Models (LLMs)… See the full description on the dataset page: https://huggingface.co/datasets/google/FACTS-grounding-public.behavior_grounding
Behaviorally Grounded User Profiles from the Wild
Open-ended, anonymized user profiles distilled from authentic social-media behavior, released with the paper
"Behaviorally Grounded User Profiles from the Wild for Personalized Alignment and Multi-Perspective Reasoning."
Persona-driven methods for personalizing LLMs typically rely on rigid synthetic personas built from a small set of
categorical attributes (age, gender, nationality). These flatten individual variation and lean on… See the full description on the dataset page: https://huggingface.co/datasets/UWaterloo/behavior_grounding.gaze-as-grounding-evidence
Gaze as Evidence for Common Grounding
Processed, window-level gaze features for Gaze as Evidence for Common
Grounding: A Cross-Corpus Analysis of MapTask and MUNDEX by Nan Li,
Albert Gatt and Massimo Poesio (MINT 2026).
Paper on arXiv ·
Hugging Face paper page ·
GitHub: data and analysis code
The dataset connects gaze measurements with reference-alignment annotations in
MapTask and retrospective understanding judgments in MUNDEX. Both corpora use
discrete behavioral gaze… See the full description on the dataset page: https://huggingface.co/datasets/chnln/gaze-as-grounding-evidence.HoT_User_Study_Data
📚 Fact-Enhanced Math Problem Dataset
Overview
This dataset contains mathematical reasoning problems where key facts are highlighted using fact tags (e.g., <fact1>, <fact2>). The dataset is designed for training and evaluating explainable AI (XAI) models, especially in fact-referencing reasoning tasks.
Each question and answer pair follows a structured format where supporting facts are explicitly referenced to improve transparency in mathematical problem-solving.… See the full description on the dataset page: https://huggingface.co/datasets/groundingauburn/HoT_User_Study_Data.retrieval_grounding
Dino Data Retrieval Grounding Preview
What This Dataset Is
This dataset is a focused retrieval-grounding preview built from four Dino Data capability slices:
search trigger detection
grounded search integration
history search trigger
history search integration
The goal is to train or inspect assistant behavior around two connected problems:
deciding when retrieval or history lookup is needed
generating answers that stay grounded to supplied evidence or prior thread… See the full description on the dataset page: https://huggingface.co/datasets/DinoDS/retrieval_grounding.facts-grounding-processed
Dataset Summary
The dataset contains prompts, context documents, and target answers that challenge models to stay grounded in provided context rather than hallucinating.Processing steps added extra features like:
prompt – consolidated instruction + user request + context
has_url_in_context – boolean flag for URLs in context
len_system, len_user, len_context – token/word length statistics
row_id – unique identifier for tracking
Dataset Structure
Splits:
train – 688… See the full description on the dataset page: https://huggingface.co/datasets/GenAIDevTOProd/facts-grounding-processed.multimodal-grounding-ooc
Multimodal Grounding of Explanations for Out-of-Context Misinformation Detection
This dataset contains the outputs, explanations, and visual grounding audits for three vision-language model configurations evaluated on out-of-context (OOC) misinformation detection:
Gemma-4-31B-It (Direct): Baseline API evaluation with minimal thinking compute.
Gemma-4-31B-It (Thinking): Deliberation API evaluation with high thinking compute (up to 4,352 tokens).
Gemma-3-27B-It (Direct):… See the full description on the dataset page: https://huggingface.co/datasets/jordansp/multimodal-grounding-ooc.expert-element-grounding_resultsarch-verifier-grounding-264
Verifier grounding study (264 graphs)
Clean reference architectures plus systematically corrupted variants (broken attention head divisibility, linear width mismatches, severed connections), each built as a real PyTorch model and run on a GPU. Every row pairs the static verifier verdict with what actually happened at runtime: whether the module constructed, whether the forward pass survived, whether training made progress, and the initial and final loss. 264 graphs, two seeds… See the full description on the dataset page: https://huggingface.co/datasets/neurarch-ai/arch-verifier-grounding-264.FACTS-grounding-public
FACTS Grounding 1.0 Public Examples
860 public FACTS Grounding examples from Google DeepMind and Google Research
FACTS Grounding is a benchmark from Google DeepMind and Google Research designed to measure the performance of AI Models on factuality and grounding.
▶ FACTS Grounding Leaderboard on Kaggle▶ Technical Report▶ Evaluation Starter Code▶ Google DeepMind Blog Post
Usage
The FACTS Grounding benchmark evaluates the ability of Large Language… See the full description on the dataset page: https://huggingface.co/datasets/rohith7820/FACTS-grounding-public.FACTS-GROUNDING-EVAL-PROMPTSFACTS Grounding is a benchmark from Google DeepMind and Google Research designed to measure the performance of AI Models on factuality and grounding. This dataset is a collection 860 examples (public set) crafted by humans for evaluating how well an AI system grounds their answers to a given context.
Each example is composed of a few parts:
A system prompt (system_instruction) which provides general instructions to the model, including to only answer the question provided based on the… See the full description on the dataset page: https://huggingface.co/datasets/avemio-digital/FACTS-GROUNDING-EVAL-PROMPTS.HoT
📚 Fact-Enhanced Math Question Dataset
Overview
This dataset contains math word, logical reasoning, question answering and reading comprehension problems with automatically reformatted questions and answers using XML tags for facts. It is designed to facilitate research in explainable AI (XAI), Human-AI interaction.
Each question is reformatted to explicitly highlight key facts using XML-style tags (<fact1>, <fact2>, etc.), and the answer explanation follows a… See the full description on the dataset page: https://huggingface.co/datasets/groundingauburn/HoT.defuser-grounding-som_resultsFACTS-GROUNDING-PUBLIC-DATASETFACTS Grounding is a benchmark from Google DeepMind and Google Research designed to measure the performance of AI Models on factuality and grounding. This dataset is a collection 860 examples (public set) crafted by humans for evaluating how well an AI system grounds their answers to a given context.
Each example is composed of a few parts:
A system prompt (system_instruction) which provides general instructions to the model, including to only answer the question provided based on the… See the full description on the dataset page: https://huggingface.co/datasets/avemio-digital/FACTS-GROUNDING-PUBLIC-DATASET.defuser-grounding-coordinates_results