We view Large Language Models as stochastic language layers in a network, where the learnable parameters are the natural language prompts at each layer. We stack two such layers, feeding the output of one layer to the next. We call the stacked architecture a Deep Language Network - DLN
Sure, we can do that once we have the links to the arxiv submission and the project page.
The instructions on how to run experiments will be included when we review the scripts.
We could add links to Arxiv, PDF, blog post like they've done here (and maybe the badges too)![image](https://github.com/microsoft/deep-language-networks/assets/660004/62b3307a-fc71-4b16-9569-64eaf0231d0f)