# Read the visualizations

Use a visual view to explore a question, then verify the underlying records.

Source: https://docs.minds.sh/docs/workspace/visualizations



The **Visualize** area offers views of relationships, agent activity, the knowledge graph, embedding space, memory, and event density. Each is a different projection of instance data.

## Choose the view for the question [#choose-the-view-for-the-question]

| Question                                      | View                             |
| --------------------------------------------- | -------------------------------- |
| How are these entities connected?             | Relationships or Knowledge Graph |
| Which agents contributed activity?            | Agent Activity                   |
| Which items are near each other by embedding? | Embedding Space                  |
| How is memory distributed?                    | Memory Map                       |
| When did activity concentrate?                | Event Density                    |

## Interpret carefully [#interpret-carefully]

Visual proximity is not always a stored relationship. In an embedding view, nearby points reflect the representation and projection used; in a graph view, an edge represents an explicit relationship. A cluster does not by itself imply causality, trust, or a recommendation.

When a pattern matters, inspect the record identifiers, filters, time window, and source query. If the instance is unavailable or a view is using fallback data, return to an authenticated API read before relying on the display.