I have a theory on what makes LLM slop tweets distinctive: low information content but high encoding complexity. See graph⬇️

The best guide for anyone to start learning networking with programming. This article will help you learn: 0. Networking protocols (TCP/UDP) 1. HTTP under the hood 2. Chunked encoding 3. Building state machines 4. Writing your own parser 5. Concurrency basics (multithreading)

My recent 7 articles on X: - KV Cache in LLMs - Paged Attention in LLMs - Causal Masking in Attention - Byte Pair Encoding in LLMs - Harness Engineering in AI - Math behind Attention - Q, K, and V - Math behind √dₖ Scaling Factor in Attention X is a knowledge sharing

Pythagorean triples, one's complement, and run-length encoding

CloudFront WAF sets a 403 interception rule for the `/actuator` path, but you can use URL encoding `/%61%63%74%75%61%74%6f%72` (That's, each character of `/actuator` is hexadecimal encoded) to bypass the WAF and directly access Spring Boot #BugHunter #BugBounty #BugBountyTips

🛳️🚢 BurgsVO: Burgs-Associated Vertex Offset Encoding #Scheme for #Detecting Rotated #Ships in #SAR Images ✍️ Mingjin Zhang et al. 🔗 https://t.co/RYgkvAC3qe

My recent 9 articles on X: - KV Cache in LLMs - Paged Attention in LLMs - Causal Masking in Attention - Byte Pair Encoding in LLMs - Harness Engineering in AI - Math behind Attention - Q, K, and V - Math behind √dₖ Scaling Factor in Attention - Math Behind Backpropagation -

AI coding agents are fast but reckless. They skip specs, tests, and security. Google engineer just open-sourced a fix. Agent Skills is a free repo that brings 19 engineering skills and 7 slash commands to any AI coding agent. It works by encoding what senior engineers

My recent 5 articles on X: - KV Cache in LLMs - Paged Attention in LLMs - Causal Masking in Attention - Byte Pair Encoding in LLMs - Harness Engineering in AI X is a knowledge sharing platform.

🚨 BREAKING: AI memory just got flipped… by a video file. No vector DB. No infra headaches. Just a single .mp4. Someone built Memvid — and it stores millions of embeddings inside a video file using encoding tricks. Sounds crazy. It works. Here’s why this is wild: → Entire

Burp Suite's "Decoder" is great for URL encoding - but there is a much faster way to do it!..

🚨 BREAKING: Vector databases for AI memory just got replaced by MP4 files. Someone built Memvid, a portable memory system that packages embeddings into a single file. It stores millions of text chunks using video encoding logic for sub-millisecond retrieval. → Replace

My recent 10 articles on X: - KV Cache in LLMs - Paged Attention in LLMs - Causal Masking in Attention - Byte Pair Encoding in LLMs - Harness Engineering in AI - Math behind Attention - Q, K, and V - Math behind √dₖ Scaling Factor in Attention - Math Behind Backpropagation -

[LLM Internals] Just published an article: Feed-Forward Networks My recent 12 articles on X: - KV Cache - Paged Attention - Causal Masking - Byte Pair Encoding - Harness Engineering - Math behind Attention - Q, K, and V - Math behind √dₖ Scaling Factor in Attention - Math

Cap 0.4.82 is here! Edit by transcript. Exports up to 97% smaller with optimized encoding option. Cursor-only ProRes export. Faster on-device transcriptions with Parakeet. Smoother cursor and zoom animations, and a lot of performance and bug fixes.

My recent 8 articles on X: - KV Cache in LLMs - Paged Attention in LLMs - Causal Masking in Attention - Byte Pair Encoding in LLMs - Harness Engineering in AI - Math behind Attention - Q, K, and V - Math behind √dₖ Scaling Factor in Attention - Math Behind Backpropagation X

Following Netflix’s latest optimization round, which introduced Film Grain Synthesis (FGS) in their AV1 encoding pipeline, the service streams some movies at bitrates as low as 200 kbps: https://t.co/hrB7D9s9TY via Fabio @Sonnati

Any local llm nerd around? I'm trying to run speculative encoding on Gemma 26B A4B. I'm newbie to running that stuff locally, got it to 200 B tokens/s on B200, I wonder if I could make it much faster?

Just published an article: Decoding Flash Attention My recent 11 articles on X: - KV Cache - Paged Attention - Causal Masking - Byte Pair Encoding - Harness Engineering - Math behind Attention - Q, K, and V - Math behind √dₖ Scaling Factor in Attention - Math Behind