Yaz claims editable AI knowledge without retraining

Yaz from Tilelli Lab is an presented as a research prototype. Its main idea is editable knowledge: facts can be added, changed, or removed without training the model again.

The intended use is custom knowledge that needs to stay current, handled more like a simple database than a fixed model. It is described as running locally on a computer’s CPU.

It is also designed to answer “I don’t know” when it is unsure instead of making up an answer. Tilelli Lab also points to two earlier releases: Atome LM, which ran on a $5 ESP32 chip and included 12 AI apps, and Tilelli LLM, which ran on a CPU and also emphasized honest uncertainty.

Key points

  • Yaz is described as an .
  • Its knowledge can be added, changed, or deleted without retraining.
  • It on a CPU, according to the release text.
  • It is designed to say “I don’t know” when unsure instead of guessing.
  • The project has a public and product page.
Read original