ExplainableML/in-context-impersonation

[NeurIPS 2023 Spotlight] In-Context Impersonation Reveals Large Language Models' Strengths and Biases

26
/ 100
Experimental

This project helps researchers and developers understand how Large Language Models (LLMs) behave when they are asked to act like different people or experts. By providing a persona as part of the input, you can see how the LLM's text output changes and if it performs better or worse on tasks. This is useful for anyone evaluating LLM performance, potential biases, or exploring their nuanced capabilities in areas like reasoning or description.

No commits in the last 6 months.

Use this if you need to systematically test how different 'personas' (like a child, a domain expert, or a specific demographic) influence an LLM's output and task performance.

Not ideal if you are looking for a tool to train new LLM models or to integrate LLM capabilities into a production application directly.

LLM evaluation AI ethics computational linguistics persona-based testing model bias detection
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 4 / 25

How are scores calculated?

Stars

22

Forks

1

Language

Python

License

MIT

Last pushed

Nov 30, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/ExplainableML/in-context-impersonation"

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