Explore

Your knowledge graph is a visual map of how everything in your memory connects.

Open Explore from the sidebar to see an interactive graph that shows your files, entities, topics, commitments, and events — and all the relationships between them. As you add more content, the graph grows and new connections form automatically.

Node types

Each item in your memory is represented as a node in the graph. Nodes are colour-coded by type:

Files & NotesDocuments, text notes, and web pages you've saved.
EntitiesPeople, organisations, tools, and named things extracted from your content.
TopicsThemes and subject areas that appear across your files.
CommitmentsAction items and tasks found in your notes.
EventsCalendar entries, deadlines, and dates.

Nodes are sized by importance — more referenced items appear larger.

How nodes connect (edge types)

The lines between nodes represent different kinds of relationships. The graph uses 11 distinct edge types to show how your knowledge links together:

Structural edges

These edges connect files to the things extracted from them.

ContainsA file contains a topic. Created when a topic is extracted from a file's content.
MentionsA file mentions an entity. Created when a person, tool, or organisation is found in a file.
Has commitmentA file has an action item or task. The weight can be adjusted by loop pressure — how urgently the commitment needs attention.
Has eventA file has a calendar entry or deadline.

Co-occurrence edges

These edges show entities and files that are related through shared content.

Co-occursTwo entities that appear together in the same files. The more files they share, the stronger the connection.
Shared entityTwo files that mention the same entity. Computed dynamically when you explore the graph.
Shared topicTwo files that cover the same topic.

Temporal edges

These edges represent time-based relationships between your memories.

Timeline spine (prev/next in time)A faint navigational line connecting consecutive files in chronological order. Always present, just for orientation.
Temporal nextA real temporal connection between files that were created close together — within a 4-hour session or 36-hour window. Stronger the closer they are in time.
Semantic bridgeConnects files across large time gaps (more than 36 hours) when they share entities or topics. This is how Woxpas finds threads of continuity across weeks or months.

How temporal weights work: Files created within 4 hours have a strong connection (0.9-1.0). Same day (4-12h) is moderate (0.7-0.9). Consecutive days drop to 0.5-0.7. Beyond 36 hours, only semantic bridges form — and only when there's shared content.

Search modes

The search bar in Explore (also accessible with Cmd/Ctrl + K) has three modes:

ExplainFor meaning-based questions. Woxpas synthesises an answer from your memory using AI, with citations back to source material. Use this when you want to understand something.
ReceiptsFor artefact retrieval. Returns the raw source content — exact quotes, documents, and files. Use this when you want to find a specific thing you saved.
AutoLet Woxpas decide. It classifies your query as meaning or artefact and picks the right mode. The resolved mode is shown after the search.

Interacting with the graph

  • Click any node to open its details — you'll see a synthesis, related items, citations, and temporal neighbours.
  • Search highlights matching nodes and their connections, then zooms the camera to fit them.
  • Pan and zoom with your mouse or trackpad to navigate.
  • The Statistics button shows graph metrics: node and edge counts by type, memory health indicators.
  • A capacity indicator at the bottom shows how much of the graph is rendered. Search always finds all nodes, even those outside the visible budget.

Memory decay

Nodes in the graph can drift toward the edges over time if you haven't interacted with them. This is the memory decay model — items you access stay central, while forgotten items move to the periphery. Clicking a node or interacting with the related content brings it back toward the centre.