IdentityLM allows you to create text from a language model that is statistically signed by you, in a way that is computationally very hard to replicate, and computationally very easy to verify.

It builds on current research around logit biasing, statistically advantaging certain tokens in a deterministic way. That research focuses on watermarking language model output to identify the model it came from. A really good paper about this idea was just awarded outstanding paper at ICML - https://arxiv.org/pdf/2301.10226.pdf.

I’ve basically taken that research and interpreted it in the context of public key cryptography. The readme of the github repo is basically small white paper describing how it works.

Why should you care?

First of all, we can fight scams and deepfakes with it which I know everyone is worried about with generative models becoming so good at simulating a politician, or writing convincing phishing emails and so on. Using speech-to-text it could also verify phone identities.

Second, it’s allows transferable identity and trust across any internet platform. You can prove who you are anywhere, and link different profiles, just with natural text.

Third, it allows extra encryption around pretty much anything. It adds an extra layer of proof to any interaction or communication.

Love to have any comments or feedback!