🕵️♀️ EvidenceSeeker Boilerplate
A code template for building customised fact-checkers
EvidenceSeeker Boilerplate is a Python code template that can be used to set up LLM-based fact-checking tools—we call them EvidenceSeeker:
- An EvidenceSeeker fact-checks a statement by searching for confirming and debunking information in a given knowledge base.
- EvidenceSeeker Boilerplate is 100%-open source and is distributed under the very permissive MIT licence.
Handling Input Ambiguities
Deal with ambiguities in the inputs being fact-checked.
Model Independence
EvidenceSeeker does not rely on a specific language model. It is up to you: Use EvidenceSeeker with your privately hosted LLMs or rely on inference service providers; use proprietary or open models.
Explainability & Transparency
Stay in control of your fact-checking workflow. An EvidenceSeeker is not a black box. Its assessments can be explained and traced back to occurrences in your knowledge base.
Easy Set-up
Use the boilerplate to set up your custom-made EvidenceSeeker without caring about technical intricacies.
Integration
Integrate EvidenceSeeker into your existing workflows and AI apps.
Domain-specific Fact-Checking
Use your knowledge base (e.g., a collection of PDFs) for fact-checking.