intuit-ai-research/DCR-consistency

DCR-Consistency: Divide-Conquer-Reasoning for Consistency Evaluation and Improvement of Large Language Models

36
/ 100
Emerging

This tool helps evaluate and improve the accuracy of text generated by large language models (LLMs) by detecting and fixing "hallucinations" or inconsistencies. It takes a reference text (ground truth) and a candidate text from an LLM, then outputs a consistency score, specific reasons for inconsistencies, and an improved version of the candidate text. This is for anyone who uses or develops applications powered by LLMs and needs to ensure their outputs are factual and reliable.

No commits in the last 6 months.

Use this if you need to objectively measure the factual consistency of LLM-generated text against a source and automatically correct identified errors.

Not ideal if your primary concern is grammar, style, or other linguistic qualities not directly related to factual consistency with a reference.

LLM evaluation content quality assurance text generation fact-checking AI model improvement
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 7 / 25
Maturity 16 / 25
Community 13 / 25

How are scores calculated?

Stars

25

Forks

4

Language

Python

License

Apache-2.0

Last pushed

May 23, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/intuit-ai-research/DCR-consistency"

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