📜🚨
Introducing TensorLens! 🔎
Our new tool for Transformer & LLM interpretability.
The problem: attention matrices are (i) a shallow view that ignores embeddings, FFNs, and values, and (ii) there are too numerous (per head & layer), which quickly becomes overwhelming.
🧵 1/6
Your intelligence cannot be measured by a number. It is defined by your willingness to learn, solve problems and try new things. You are more than just a number. Develop your skills. Share your brilliant ideas. Your skills are more valuable than your grades. 🧠