<aside> 📖

https://github.com/ml5js/Intro-ML-Arts-IMA-F24/tree/main/07-llms

</aside>

Language models can only write ransom notes (Allison Parrish - 2024)

Language models can only write ransom notes

The Foundation Model Transparency Index

Foundation Model Transparency Index

AI Model Experimentation - Ollama

For this assignment I decided to stress-test 3 LLMs from major tech corporations. These included:

Llama3.2 - 3.2b

Frame 2.png

Gemma2.0 - 9.2b

Frame 1.png

Phi3.5 - 3.8b

Frame 3.png

As for the questions, I tried to think of different questions which could “trick” these models.

There were a couple assumptions I wanted to test: