Category: Uncategorized
-
a ranked annotated bibliography – LLM-to-LLM communication in latent space
1) Communicating Activations Between Language Model Agents — Vignav Ramesh, Kenneth Li (2025) Link: arXiv page. (arXiv) Expanded abstract summary This paper is one of the clearest direct proposals for replacing text-based inter-agent communication with an internal, continuous alternative. The authors begin from a simple observation: when LLM agents “talk” to one another in natural…
-
Latent Packets and Machine Telepathy: A Frank-said / GPT-said Dialogue on Semantic Geometry and the Future of AI Communication
Frank said: I keep coming back to the same intuition: if LLMs do not really “think” in English, but instead move through some hidden semantic geometry, then ordinary language is just the outer shell. It is the human-readable exhaust. The real action is happening underneath, in latent space. So I want to know what it…
-
From Latent Space to Living Mind: A Frank-said / Claude-said Dialogue on Whether Machines Can Mean Anything at All
Frank said: What fascinates me here is that this may not just be a new trick for sending more data. It may be the beginning of a different idea of communication altogether. Not merely faster transmission, not merely more channels, but a shift from communication as symbol delivery to communication as structured geometry. If orbital…
-
From Twisted Light to Latent Space: A Frank-said / GPT-said Dialogue on the Future of Communication
Frank said: What fascinates me here is that this may not just be a new trick for sending more data. It may be the beginning of a different idea of communication altogether. Not merely faster transmission, not merely more channels, but a shift from communication as symbol delivery to communication as structured geometry. If orbital…
-
From Meaning to Geometry: A Frank-Said / GPT-Said Dialogue on Embeddings, Weights, and Inference
Frank said:When an LLM creates an embedding it converts real world information into a mathematical representation of the information and that mathematical representation is a multidimensional vector that capture the meaning of the information in the vectors direction and size. Each multidimensional vector that the LLM creates has a relationship to every other multidimensional vector…
-
Title: How Geometry Becomes Behavior – frank said/gpt said
Frank said: I want to go deeper into this because it feels like one of those hidden mechanisms at the heart of AI. We talk about embeddings as these multidimensional vectors, and we talk about weights as the substance of the neural network, but the relation between the two still feels slippery. How does something…