This is Version 1.0 of the neural memory network brain, comprising 180,000 neurons and nearly 9 million synapses; it primarily encompasses an interconnected knowledge graph spanning fields such as mathematics, physics, and AI. The graph is structured around connections based on formal mathematical proofs, forming a small-world network.

Statistics

Metric	Value
Neurons (concepts)	181,462
Synaptic connections	8,440,886
Average degree	~46.5
Neurotransmitter types	6 (GABA, ACh, Glu, DA, 5-HT, Oct)

File manifest

File	Description
`neurons.json`	Full neuron array — id, content, membrane potential `v`, conductance `g`, neurotransmitter, type, spike count, importance, tags, source papers, created timestamp, and outgoing/incoming connection weights.
`connectivity.parquet`	Sparse synaptic adjacency (source, target, weight).
`embeddings.npy`	Semantic embedding vectors for each neuron.
`embedding_ids.json`	Neuron-id → embedding-row index mapping.
`evolution_state.json`	Learning/evolution bookkeeping state.
`meta.json`	Snapshot metadata (saved timestamp, neuron count, connection count).
`.p0_dubrovin_checkpoint.json`	Incremental feeding checkpoint.

License

Unspecified (other). Contact the owner for usage terms.

Downloads last month: 7

Total file size:

1.27 GB