google-research/fool-me-twice
Game code and data for Fool Me Twice: Entailment from Wikipedia Gamification https://arxiv.org/abs/2104.04725
This project offers a gamified approach to help players identify whether statements are supported or refuted by evidence found on Wikipedia pages. Players read a claim, examine provided Wikipedia text, and then decide if the claim is entailed or not. It also allows players to create their own challenging statements based on Wikipedia content. This is designed for researchers, educators, or anyone interested in collecting high-quality, human-annotated datasets for natural language understanding and fact-checking tasks.
Use this if you need to create or collect a dataset of human-verified statements and their supporting/refuting evidence from Wikipedia, particularly for natural language inference research or educational purposes.
Not ideal if you are looking for a pre-built, production-ready game to host for a general audience without significant development and setup.
Stars
25
Forks
8
Language
JavaScript
License
Apache-2.0
Category
Last pushed
Mar 14, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/google-research/fool-me-twice"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
mihail911/fake-news
Building a fake news detector from initial ideation to model deployment
FakeNewsChallenge/fnc-1-baseline
A baseline implementation for FNC-1
palewire/storysniffer
Inspect a URL and estimate if it contains a news story
IKMLab/CFEVER-data
AAAI-24 CFEVER: A Chinese Fact Extraction and VERification Dataset
pmacinec/fake-news-datasets
This repository contains list of available fake news datasets for data mining.