distributed neural network

• make visualization of soft3 architecture
• here we present new architecture of distributed neural network
• layers
  • input: ask
    • soul: define default processing rules for processing
    • will regulated by $V: limits cybergraph bandwidth
    • neural: expressive semantic language for cybergraph
    • attention regulated by $A: affects probabilities of random walk
    • random walk measurements: get probabilities on nodes
    • standard inference: compute on gpu truthful order of particles in context
    • dynamic names: ability to map static names to js and wasm code
    • cyber-cw set of semantic cosmwasm progs for learning during execution
    • processing ordered list of particles by llm: local or cloud
    • motivation driven by $O: allow to cover cost base of learning through learning rewards
  • output: answer
• that is what chatgpt gave me to a query of creating diagram of proposed architecture
  • 
• hopefully proposed architecture will be able to demonstrate better results
• related reading
  • knowledge graphs and llms
• discover all concepts