IAAR-Shanghai/UHGEval

[ACL 2024] User-friendly evaluation framework: Eval Suite & Benchmarks: UHGEval, HaluEval, HalluQA, etc.

40
/ 100
Emerging

This tool helps researchers, product managers, and AI safety engineers evaluate how often Chinese Large Language Models (LLMs) generate incorrect or made-up information. It takes an LLM and a prompt as input, and then uses various methods to measure the 'hallucination rate' of the generated Chinese text. The output helps users understand how reliable specific Chinese LLMs are for tasks requiring factual accuracy.

180 stars. No commits in the last 6 months.

Use this if you need a comprehensive and user-friendly way to measure the factual accuracy and hallucination tendencies of Chinese LLMs across different benchmarks.

Not ideal if your primary concern is evaluating LLMs for English content or if you require fine-grained analysis beyond hallucination, such as toxicity or bias.

LLM evaluation AI safety Chinese NLP hallucination detection content verification
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 12 / 25

How are scores calculated?

Stars

180

Forks

13

Language

Python

License

Apache-2.0

Last pushed

Jun 07, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/IAAR-Shanghai/UHGEval"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.