Skip to main content
Glama
Sign In
Neuronpedia MCP Server

Neuronpedia MCP Server

by manncodes
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript

Neuronpedia MCP Server

A Model Context Protocol (MCP) server that provides access to the Neuronpedia API for AI model interpretability and feature analysis.

Features

  • Explanations: Generate and search explanations for AI model features

  • Activations: Analyze feature activations on specific text inputs

  • Steering: Control model generation using specific features

  • Vectors: Create and manage custom steering vectors

Related MCP server: Orchestra MCP Server

Installation

npm install
npm run build

Configuration

Set your Neuronpedia API key as an environment variable:

export NEURONPEDIA_API_KEY=your_api_key_here

You can obtain your API key by logging in at neuronpedia.org and visiting your account page.

Usage

Running the server

npm start

Available Tools

  1. generate_explanation - Generate explanations for model features

  2. search_explanations - Search existing explanations

  3. get_activations - Get feature activations for text

  4. search_top_features - Find top activating features

  5. steer_generation - Steer model generation with features

  6. create_vector - Create custom steering vectors

  7. list_vectors - List all vectors

  8. get_vector - Get specific vector

  9. delete_vector - Delete a vector

MCP Client Configuration

Add this to your MCP client configuration:

{
  "mcpServers": {
    "neuronpedia": {
      "command": "node",
      "args": ["/path/to/neuronpedia-mcp/dist/index.js"],
      "env": {
        "NEURONPEDIA_API_KEY": "your_api_key"
      }
    }
  }
}

Development

npm run dev

License

MIT

Available Tools

9 tools
create_vectorB

Create a custom vector for steering

Inputs: name (required), values (required), steeringStrength.

delete_vectorB

Delete a vector by ID

Inputs: vectorId (required).

generate_explanationC

Generate an explanation for a specific feature in an AI model

Inputs: layer (required), model (required), feature (required).

get_activationsB

Get activation values for a specific feature on given text

Inputs: text (required), layer (required), model (required), feature (required).

get_vectorB

Get a specific vector by ID

Inputs: vectorId (required).

list_vectorsA

List all custom vectors

Inputs: no named parameters.

search_explanationsB

Search for explanations across models and layers

Inputs: query (required), layer, model.

search_top_featuresC

Find the top activating features for given text

Inputs: text (required), layer (required), model (required), topK.

steer_generationC

Steer model generation using a specific feature

Inputs: layer (required), model (required), prompt (required), feature (required), +2 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A3.7/5.0

    Scored across 9 tools

    Disambiguation5/5

    Each tool targets a distinct operation: explanation generation/search, activation retrieval, feature search, steering, and vector management. No two tools have overlapping purposes, ensuring clear disambiguation.

    Naming Consistency5/5

    All tool names follow a consistent snake_case verb_noun pattern (e.g., generate_explanation, list_vectors, delete_vector), making the set predictable and easy to navigate.

    Tool Count5/5

    With 9 tools, the set is well-scoped for a focused domain like neuronpedia. It covers essential operations without being too sparse or overwhelming.

    Completeness5/5

    The tool surface provides a complete workflow: explanation generation and search, activation analysis, feature search, steering via vectors, and full CRUD for custom vectors. No obvious gaps.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Resources

    Related MCP Connectors

    Related MCP Servers

    • Jina AI MCP Server

      zk-armor
      A
      license
      Not graded
      quality
      D
      maintenance
      Provides access to Jina AI's Search Foundation APIs for embeddings, web search, content extraction, reranking, classification, and semantic text segmentation.
      18 npm
      MIT

    • synapse-mcp

      marvellousz
      F
      license
      Not graded
      quality
      D
      maintenance
      Connects AI clients to a Synapse backend API, enabling memory operations (search, list, create, get) and question answering through natural language.
      -

    • neurograph

      InsideProxy
      F
      license
      A
      quality
      B
      maintenance
      Exposes NeuroGraph's scientific API to AI models as MCP tools, enabling them to query, analyze, and visualize brain networks — regions, tracts, connectivity, literature, and cross-species homologies — drawn from a portable SSD-based neuroimaging data library. Lets models read real atlas data, load datasets, compute graph metrics, and produce visualization scene descriptors without acting as the scientific engine itself.
      12
      3
      -