- Published on
GPT: from the decoder-only simplification of Transformer through pre-training, scaling laws, and RLHF alignment, tracing the evolution from GPT-1 to GPT-5.
Welcome
A collection of notes and technical deep dives on transformers, large language models, and AI engineering.
- Published on
A deep-dive into the Transformer architecture: tokenization, positional encoding, multi-head attention, and encoder-decoder design explained with visuals.