Blog: Jack Kowalski

HMFoDG as an audit for UMAP and UPGMA. Hierarchical clustering on a subset that is clusterable in the first place.

HMFoDG is a tool that tests whether you are even allowed to run UMAP and UPGMA on one piece, or whether you must split into subsets - because neither of those tools can detect the absence of a common trunk / the absence of a shared quasinorm.. Read more…

Detecting a false common space, why L2 is not an ontology.

Mapping mixed-norm aggregations - notes from a long conversation about a shared vocabulary with biotech. Read more…

Trace conclusiveness of contextual graphs and knowledge.

Each argument has its own map of routes that changes depending on the order in which and with what you connect it. Read more…

HMFoDG Implementation - Analytical Curiosity for RAG Auditing.

In the Hierarchical Metric Flow on Data Graphs (HMFoDG) implementation, an interesting analytical feature emerged. It enables auditing of RAG systems (and embeddings or any vectorized data - biotech, medical, finance, physics, etc.) by tracing the selected associations in a deterministic way (without heuristics). This means deterministically tracking what the transformer actually used during the reinforcement process that caused one token to lie close to another. These are very concrete references (e.g., author, specific tokens from the text that in sufficient quantity allow source identification). Read more…

Back to top