The Parameter Explosion in Multi-Scale Neuromodulation Models

{
  "title": "The Parameter Explosion in Multi-Scale Neuromodulation Models",
  "metric": "Order of magnitude of free parameters",
  "papers": [
    {
      "doi": "aggregate",
      "value": "~5-10",
      "condition": "Gain modulation models",
      "model_scale": "Single neuron",
      "value_source_sentence": "Single-neuron gain modulation models typically have 5-10 free parameters describing receptor-specific gain functions"
    },
    {
      "doi": "aggregate",
      "value": "~50-200",
      "condition": "Microcircuit neuromodulation models",
      "model_scale": "Local circuit (~100 neurons)",
      "value_source_sentence": "Circuit-level models (microcircuit with multiple cell types and neuromodulator receptors) typically have 50-200 parameters"
    },
    {
      "doi": "aggregate",
      "value": "~1000-10000",
      "condition": "Brain-wide mean-field models",
      "model_scale": "Whole-brain (~100 regions)",
      "value_source_sentence": "Whole-brain mean-field models with region-specific neuromodulatory coupling have ~1000-10000 parameters across regions x modulators x receptor types"
    }
  ],
  "n_analyzed": "Estimated from model descriptions across literature",
  "description": "Comparison of approximate parameter counts across different scales of neuromodulation models, from single-neuron to whole-brain",
  "n_definition": "Approximate count of free parameters in published models",
  "scope_region": "Multiple scales from neuron to whole brain",
  "comparison_id": "fig_sec11_parameter_explosion",
  "taxonomic_level": "Not applicable",
  "scope_population": "Published computational models",
  "homogeneity_check": "Low homogeneity: parameter counts are estimates since many papers don't explicitly report total free parameters. 'Free parameters' is ambiguous - some include fixed biophysical constants, others only count fitted parameters. These are order-of-magnitude estimates suitable only for illustrating the scaling challenge."
}

[]

{
  "mode": "public_source_pointer_with_short_context",
  "notes": [
    "Local review repositories are read-only inputs.",
    "SciDEX stores paper metadata, structured evidence, file pointers, and short citation contexts; it does not copy full review prose."
  ],
  "source_commit_sha": "95db5c630fe54e183d9c452cf826ce502d4a872d",
  "source_repository_url": "https://github.com/AllenNeuralDynamics/ComputationalReviewNeuromodulation"
}