Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.memorybox.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
The Engine Room

The Advanced tab provides granular control over the MemoryBox backend. This is where you configure the logic of your Synaptic Archive and choose exactly which models are available in your Command Center.
Advanced settings overview showing engine calibration

​
👥 AI Expert Selection (Platform Models)

This section allows you to curate your roster of AI Experts. Rather than being overwhelmed by every model on the market, you can choose exactly which ones appear in your Command Center’s model picker.
Platform Models selection menu
  • Curating Your Team: Click the down-caret under Platform Models to reveal the full list of supported LLMs (GPT-4o, Claude 3.5, Gemini, etc.).
  • Active Roster: Toggle the switch next to each model to “Show” or “Hide” it. This ensures that your workspace remains clean and only features the experts you trust for your specific workflow.
  • API Configuration: If you are using personal API keys, this is where you ensure the connection is active for each selected platform.

​
⚙️ Synaptic Archive Calibration

The bottom half of the Advanced pane controls how MemoryBox processes and retrieves your information. These settings define the “intelligence” of your memory grounding.

​
Memory Retrieval

  • Embedding Model: Choose the underlying model used to “index” your thoughts.
  • Rerank Mode: Select the logic used to prioritize relevant memories during a search.
  • Top K: Define how many “memory engrams” the AI should pull for each prompt. A higher number provides more context but may increase processing time.

​
Technical Performance

  • Chunk Max Chars: Control the size of the “information bites” MemoryBox uses when storing your documents.
  • Timeout Settings: Adjust how long the system waits for an expert to respond before a “Rerank Timeout” or “Connection Timeout” occurs.

​
🧹 Maintenance & Integrity

At the bottom of the pane, you have tools to manage the health of your local database.
  • Clear Local Database: Use this with caution to wipe your local Synaptic Archive and start fresh.
  • Clear User Config: Resets all your theme, language, and model preferences to the factory default.

​
💾 Committing Your Configuration

Because the Advanced settings directly impact the Synaptic Archive engine, changes must be explicitly applied or managed to take effect. Use the controls in the bottom-right corner of the pane to manage your session state.
  • Apply Config: Click this to commit your adjustments to the local database and update your active AI Experts.
  • Load Current Config: Refreshes the view to show the settings currently running in the engine (useful if you have made manual changes elsewhere).
  • Reset to Loaded: Discards any pending changes in the current view and reverts the UI to match the last successfully applied configuration.
Engine Persistence: Simply closing the window will not “Apply” your changes. To ensure your model selections and memory logic are updated for your next ideation session, you must click Apply Config before exiting.