馃憢 Hi, all!
I am Elliot Shichao Song, a PhD candidate. My current research focuses on foundation models. I blog about my research and life.
)](https://ki-seki.github.io/posts/261025-training-data-retrieval/assets/semantic-islands-cover.png)
A Billion-Scale Retrieval System for Text-to-Image Training Data
Training a text-to-image model from scratch means finding, inspecting and filtering samples across billions of image鈥搕ext pairs. Using a retrieval system over 1.4B samples as an example, this post covers data storage and management, distributed embedding extraction, building and querying an IVF-PQ index, and online retrieval and cluster browsing.
)](https://ki-seki.github.io/posts/260531-lilian-rl-overview/RL_illustration_new.png)
Fully Annotated Guide to "A (Long) Peek into Reinforcement Learning"
This is a fully annotated guide to Lilian Weng鈥檚 post A (Long) Peek into Reinforcement Learning.
)](https://ki-seki.github.io/posts/260430-multi-armed-bandit/multi-armed-bandit.png)
Fully Annotated Guide to "The Multi-Armed Bandit Problem and Its Solutions"
The multi-armed bandit problem is a classic exploration鈥揺xploitation dilemma in reinforcement learning. Lilian Weng鈥檚 post is an excellent introduction, but some mathematical details and motivations can be cryptic. This article annotates it with step-by-step explanations and supplementary notes.
)](https://ki-seki.github.io/posts/260322-ouroboros/process.png)
Product Requirements Document of Ouroboros
An agentic DOM workspace where an LLM has full read/write/delete privileges over its own source code and visual interface.
)](https://ki-seki.github.io/posts/260314-ai-economics/decouple.png)
Everything is Fleeting: A Roadmap to the Post-Labor AI Economy
A speculative roadmap of how AI will dismantle the labor market and what comes after.