Essays · Notes · Technical Deep Dives
Articles
10
Updated
September 27, 2026
September 27, 2026 · 58 min read · Backend Engineering
Pulling HTTP apart from the wire up: exploring statelessness, request methods, headers, idempotency, CORS, caching, content negotiation, compression, persistent connections, and the TLS handshake behind HTTPS.
June 24, 2026 · 59 min read · Deep Learning
A ground-up walkthrough of the original encoder–decoder Transformer—from masked self-attention and cross-attention through the full encoder and decoder stacks, training-time parallelization, and autoregressive inference.
June 20, 2026 · 21 min read · Deep Learning
A first-principles explanation of positional encoding in Transformers—from why self-attention needs order information, through sinusoidal encodings at multiple frequencies, to the full PE formula and a worked example.
June 18, 2026 · 58 min read · Deep Learning
A first-principles walkthrough of self-attention—from static embeddings and manual context weighting to Query, Key, and Value projections, scaled dot-product attention, and the full matrix formulation behind modern Transformers.
June 17, 2026 · 28 min read · Deep Learning
A detailed exploration of Bahdanau and Luong Attention, including alignment scores, additive vs multiplicative attention, context vectors, attentional hidden states, and the mathematical foundations behind sequence-to-sequence attention mechanisms.
June 16, 2026 · 12 min read · Deep Learning
An intuition-first explanation of the attention mechanism in sequence-to-sequence models, covering context vectors, alignment scores, softmax, and dynamic attention in neural machine translation.
June 15, 2026 · 16 min read · Deep Learning
A walkthrough of the sequence-to-sequence encoder-decoder model, from context vectors and teacher forcing to the improvements in the original paper.
June 4, 2026 · 10 min read · Deep Learning
How pre-trained CNNs, feature extraction, and fine-tuning let you adapt deep models without training from scratch.
June 3, 2026 · 12 min read · GenAI
Exploring how retrieval, embeddings, and generation work together to ground LLM responses.
August 15, 2025 · 6 min read · Blockchain
An exploration of the processes and machinery that power Bitcoin mining.
