Julep Archive

company

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

diwank updated a dataset about 1 year ago

julep-ai-archive/storycommonsense-chatml

diwank updated a dataset over 1 year ago

julep-ai-archive/samantha-dataset-v4-instructonly-prompt_response-2048

diwank updated a model over 1 year ago

julep-ai-archive/samantha-1-tokenizer

View all activity

julep-ai-archive's activity

diwank

posted an update 6 months ago

Post

2190

Just published "CryptGPT: A Simple Approach to Privacy-Preserving Language Models Using the Vigenere Cipher".

https://huggingface.co/blog/diwank/cryptgpt-part1

tl;dr - we pretrained a gpt-2 tokenizer and model from scratch on a dataset encrypted with Vigenere cipher and it performs as well as regular gpt-2. Except in order to use it, you need to know the encryption key.

links:
https://github.com/creatorrr/cryptgpt
diwank/cryptgpt
diwank/cryptgpt-large

2 replies

diwank

posted an update 8 months ago

Post

1683

Really excited to read about Kolmogorov Arnold Networks as a novel alternatives to Multi Layer Perceptrons.

Excerpt:
> Kolmogorov-Arnold Networks (KANs) are promising alternatives of Multi-Layer Perceptrons (MLPs). KANs have strong mathematical foundations just like MLPs: MLPs are based on the universal approximation theorem, while KANs are based on Kolmogorov-Arnold representation theorem. KANs and MLPs are dual: KANs have activation functions on edges, while MLPs have activation functions on nodes. This simple change makes KANs better (sometimes much better!) than MLPs in terms of both model accuracy and interpretability.

https://github.com/KindXiaoming/pykan