When a dangerous model is deployed, it will pose misalignment and misuse risks. Even before dangerous models exist, deploying models on dangerous paths can accelerate and diffuse progress toward dangerous models.| ailabwatch.org
Note: This assessment was conducted in June 2023. The EU AI Act and the practices of companies have considerably evolved since then: the results below are only reflective of the status quo as of June 2023.| crfm.stanford.edu
We introduce Alpaca 7B, a model fine-tuned from the LLaMA 7B model on 52K| crfm.stanford.edu