datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
CultureBankbenchmarking-cultures-25
Benchmarking-Cultures-25 Dataset
This dataset accompanies the Unsteady Metrics and Benchmarking Cultures of AI Model Builders paper submitted to FAccT 2026 by Stefan Baack, Christo Buschek and Maty Bohacek.
The dataset contains the following parts:
core: The curated Benchmarking-Cultures-25 dataset.
derived: Datasets that were derived from the core dataset and informed the FAccT submission.
figures: Figures generated from derived data and used in the paper.
docs: Data dictionaries… See the full description on the dataset page: https://huggingface.co/datasets/matybohacek/benchmarking-cultures-25.employment_transportation_tourism_culture_suomipact-culture-personalization
PACT: Personal-Preference and Cultural-Norm Trade-off
This dataset accompanies Whose Norms? Disentangling Cultural and Personal Alignment in Large Language Models.
Hugging Face repository: Angana192/pact-culture-personalization
PACT contains social scenarios where a cultural expectation and an actor's personal preference are both plausible but may conflict. This release contains only the benchmark scenario instances: no model outputs, no model results, no trace-analysis tables… See the full description on the dataset page: https://huggingface.co/datasets/MichiganNLP/pact-culture-personalization.CultureVidBench
CultureVidBench Dataset
Dataset Description
CultureVidBench is a benchmark for evaluating cultural understanding in text-to-video (T2V) generation.
The benchmark contains 1,000 prompts covering:
12 countries
6 continents
8 cultural regions
3 cultural categories
14 cultural aspects
Project Page: https://hanxjing.github.io/CultureVidBench/
Cultural Categories
Category
Aspects
Number of Prompts
Material Culture
Food, Clothing, Architecture… See the full description on the dataset page: https://huggingface.co/datasets/XianjingHan/CultureVidBench.philosophy-culture-translations-html-csv
AI-Culture Philosophy and Culture Translations CSV + HTML Corpus
The corpus contains an exceptionally diverse range of cultural, philosophical, and literary texts, available in 12 major languages. Among other topics, there is extensive engagement with the ethics and aesthetics of artificial intelligence and its cultural and philosophical implications, as well as connections between AI and philosophy of language and philosophy of mind.
This project is maintained by a non-profit… See the full description on the dataset page: https://huggingface.co/datasets/AI-Culture-Commons/philosophy-culture-translations-html-csv.clinical-antibiotic-blood-culture-timing-coherence-risk-v0.1What this repo is for
Detect when blood culture collection
and antibiotic start timing
fall out of alignment
before
microbiology signal is lost
and treatment becomes blind.
clinical-antibiotic-order-culture-update-coherence-risk-v0.1What this repo is for
This dataset tests whether a model can detect coupling failures between starting antibiotics and updating therapy after microbiology results.
It measures alignment between
empirical antibiotic initiation
and
review and adjustment after culture results.
It predicts
unnecessary broad-spectrum continuation
missed de-escalation after negative cultures
delayed targeting after positive cultures
antimicrobial resistance risk signals.
You use it for
stewardship workflow auditing… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-antibiotic-order-culture-update-coherence-risk-v0.1.CultureInspiredPrompts
CultureInspiredPrompts
tags: CulturalSensitivity, ImageClassification, PromptGeneration
Note: This is an AI-generated dataset so its content may be inaccurate or false
Dataset Description: The "CultureInspiredPrompts" dataset comprises text prompts designed to facilitate the generation of culturally sensitive and nuanced images through machine learning models. The dataset is intended for researchers and developers focusing on improving cultural awareness in AI-generated content.… See the full description on the dataset page: https://huggingface.co/datasets/infinite-dataset-hub/CultureInspiredPrompts.clinical-antibiotic-culture-result-coherence-risk-v0.1What this repo is for
Detect when
antibiotic choice
and
actual culture results
fall out of alignment
before
treatment failure
or resistance escalation.
clinical-antibiotic-prescription-culture-result-coherence-risk-v0.1What this repo is for
Detect when
antibiotics start
but microbiology feedback
does not change practice
Common breaks
no cultures before antibiotics
positive culture with no review
resistant result with no switch
negative culture with no stop
plan stated but not executed
Examples you can use
sepsis treated but no blood cultures drawn
ESBL result but piperacillin-tazo continued
cultures negative but broad spectrum continues
You use it to flag
harm risk
resistance risk
stewardship failure
CultureBankclinical-antibiotic-order-culture-result-deescalation-coherence-risk-v0.1What this repo is for
Detect when
empiric antibiotics start
but cultures do not drive
stop or narrowing
Common breaks
positive culture but no adjustment
negative culture but therapy continues
review missing
late narrowing
Examples you can use
broad spectrum continued despite negative cultures
positive culture but wrong antibiotic choice
culture result not reviewed on weekend
You use it to flag
stewardship failure risk
resistance pressure risk
Termes_pedagogiques_artistiques_et_culturels
[!NOTE]
Dataset origin: https://www.eurotermbank.com/collections/938
turismo-e-cucina-italianaCIVIC_culture
CIVIC-Culture Calibration Benchmark
Dataset Summary
The CIVIC-Culture Calibration Benchmark is a culturally grounded diagnostic dataset designed to evaluate how language models reason about normative social, ethical, and epistemic questions across cultures.
The dataset presents a set of culturally diagnostic prompts paired with region-specific normative completions, enabling systematic analysis of cultural alignment, value sensitivity, and cross-cultural reasoning… See the full description on the dataset page: https://huggingface.co/datasets/ResearchUser/CIVIC_culture.