MASTASCUSA HOLDINGS
Free Preview No payment · No account · Live data

Mastascusa Holdings · A free tool

The Knowledge Base

A live, browsable snapshot of an AI research knowledge base — continuously crawled across arXiv (cs.AI, cs.LG, cs.CL), the major frontier labs (OpenAI, Anthropic, Google DeepMind, Google AI), Hugging Face, and Hacker News. Items are embedded with a sentence-transformer model, projected into two dimensions, and rendered as a navigable semantic map.

Illustrative preview of an internal system. Click any point on the map to inspect the underlying record.

Items tracked

4,232

Sources

9

Embedded

4,232

Window

2026-04-29 → 2026-05-01

Semantic map

Each point is one record — paper, model, or announcement.

Position is derived from semantic similarity: items that share vocabulary, citations, and ideas cluster together. Color encodes source. Click any point to inspect the record; search and source-filter narrow the view.

Semantic map of the AI research knowledge base An interactive scatterplot of 4,232 records. Each point represents one paper, model release, or announcement, positioned by semantic similarity and colored by source. Drag to pan, scroll or pinch to zoom, click any point to inspect. A keyboard-friendly list view follows below.
drag · scroll = zoom · R = reset · 0 = fit · click = inspect

Click a source to toggle its dots on the map

Inspector

Selected record

Selecting a point on the map populates this panel with the record's title, source, publication date, summary, and a link to the original.

No selection yet.

Recent additions

A representative sample, one per source.

For your organization

Private knowledge bases as a service.

The same pipeline — ingestion, embedding, clustering, retrieval — built against your own sources: research literature, contracts, customer transcripts, internal documents. Delivered with a semantic map, a search interface, and an automated daily digest of what materially changed.