You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
This research is betting against llms, to be the true path to agi, throughout our work we have empirical results demonstrating our models have cognitive capabilities such as online learning, adaptive computing , president memory native to model. Llms architectures fundamentally lack the ability to have this capabilities.
As of now we have proven our capabilities for models up to 60M parameters, our main goal right now is to scale the architecture to 500M-1B parameters , and we cannot achieve this without funding , Our models require massive compute 3-5x what transformers need for same parameter count, so our one way to achieve our goals is through funding, our models have demonstrated and achieved its capabilities with empirical results at small scale.
This funding will all be targeted towards compute so we are able to scale our model.
Its only me, I am an 18yr old , from Cairo/Egypt, I work as independent researcher, I have made a research before enabling native memory to LLms through designing memory augmented transformer architecture.
The only thing that will make this project fail is lack of funding, our results is already validated at small scale and empirically proven.
I haven't raised any money, I was always self-funded, but now I am in need of funding to be able to continue