Latent Space: The New Map of Human Creativity
Latent space is a multidimensional map of human knowledge that serves as a revolutionary new medium for exploring and creating within the realm of the possible.
Vector Embeddings
The use of vector embeddings to represent text, images, or other data as dense numerical vectors for semantic similarity search, clustering, caching, and retrieval in AI systems.
Reading List
Epicure: Mapping the Culinary and Chemical Geometry of Food
438
Epicure uses a massive multilingual recipe dataset and chemical graphs to create specialized ingredient embeddings that bridge the gap between culinary practice and food chemistry.
YourMemory: Biologically-Inspired Persistent AI Memory
YourMemory provides AI agents with a persistent, biologically-inspired memory layer that uses decay and hybrid retrieval to retain important information across sessions.
SentrySearch: Semantic Video Search for Dashcams
428
SentrySearch enables semantic natural language search and automatic clipping of dashcam footage using Gemini's multimodal video embeddings.
Canonicalizing LLM Labels with Embeddings and DSU
318
Use embeddings + vector search + DSU clustering to canonicalize LLM-generated labels, yielding consistent, cheaper, and faster classification at scale.
Why Embeddings Got Bigger—and Where Efficiency Pulls Them Next
113
Embeddings got bigger with Transformers and APIs, but new efficiency techniques and infrastructure mean the future is about smarter—not just larger—dimensions.
The Dimensional Ceiling of Single-Vector Embedding Retrieval
151
Embedding-based retrieval hits a hard top-k capacity ceiling set by embedding dimension, and real systems already run into it.