Upload Nauro H01 cortex network — connectome, model, and visualizations

README.md

@@ -12,112 +12,60 @@ tags:
   - human-brain
   - temporal-cortex
   - brain-inspired
+  - spiking-neural-network
+  - reservoir-computing
 pipeline_tag: other
 datasets:
   - google/h01-release
 ---
 
-# Nauro — H01 Human Cortex Neural Network
+# Nauro — H01 Human Cortex Connectome (Full)
 
-A **neural network** whose units and connections are derived *exactly* from a
+The **complete** neuron-to-neuron connectivity matrix extracted from a
 nanometer-resolution reconstruction of **human temporal cortex**
 ([H01 dataset](https://h01-release.storage.googleapis.com/data.html),
 Google/Harvard/Lichtman Lab).
 
-## What is this?
+Built from all 166 Avro synapse shards (~32 GB raw data), filtered at
+≥0.50 confidence. This is the full connectome — no spatial cropping.
 
-This repository contains a fully initialized PyTorch `nn.Module` with
-**2,106 units** and **3,041 connections** arranged in the
-exact same pattern as real neurons and synapses in a full of 1 mm³ sample of
-a human brain.
-
-Every unit occupies its real 3-D soma position.  Every connection carries its
-measured synaptic strength.  Dale's law is enforced: excitatory neurons
-contribute only positive current, inhibitory neurons only negative.
+## Summary
 
 | Property | Value |
 |----------|-------|
-| Neurons | 2,106 |
-| Excitatory | 1,466 |
-| Inhibitory | 475 |
-| Synapses (non-zero) | 3,041 |
-| Connectivity density | 0.07% |
-| External inputs | 2,576,814 |
-| Volume | full of 1 mm³ |
-
-## Architecture
-
-| Parameter | Value |
-|-----------|-------|
-| Network type | Recurrent (connectome topology) |
-| Activation | relu |
-| Inhibitory gain | ×4 |
-| Weight scaling | 1 / mean_in_degree |
-| Dale's law | Enforced (frozen sign mask) |
+| Neurons | 16,087 |
+| Excitatory | 10,531 (65.4%) |
+| Inhibitory | 4,688 (29.1%) |
+| Non-zero connections | 76,903 |
+| Raw edges (pre-aggregation) | 116,611 |
+| Connectivity density | 0.030% |
+| Mean in-degree | 4.8 |
+| Max in-degree | 70 |
+| External inputs (total) | 27,022,313 |
+| Volume | Full 1 mm³ |
+| Cortical layers | L1–L6 + white matter |
+| Build | All 166 GCS shards, min_confidence=0.50 |
 
 ## Connectivity by cortical layer
 
-| Layer | Neurons | Exc | Inh | Connections | Density |
-|-------|---------|-----|-----|-------------|---------|
-| Layer 1 | 48 | 18 | 3 | 0 | 0.0000% |
-| Layer 2 | 120 | 66 | 43 | 51 | 0.3542% |
-| Layer 3 | 320 | 200 | 106 | 420 | 0.4102% |
-| Layer 4 | 1,002 | 748 | 196 | 1,367 | 0.1362% |
-| Layer 5 | 441 | 314 | 100 | 385 | 0.1980% |
-| Layer 6 | 106 | 80 | 13 | 112 | 0.9968% |
-| White matter | 3 | 0 | 0 | 1 | 11.1111% |
-| unclassified | 66 | 40 | 14 | 8 | 0.1837% |
+| Layer | Neurons | Exc | Inh | Internal connections | Density |
+|-------|---------|-----|-----|---------------------|---------|
+| Layer 1 | 827 | 85 | 586 | 55 | 0.008% |
+| Layer 2 | 4,656 | 2,952 | 1,594 | 21,845 | 0.101% |
+| Layer 3 | 2,692 | 1,673 | 965 | 11,018 | 0.152% |
+| Layer 4 | 3,440 | 2,622 | 688 | 8,748 | 0.074% |
+| Layer 5 | 2,313 | 1,665 | 505 | 6,252 | 0.117% |
+| Layer 6 | 1,077 | 906 | 128 | 4,419 | 0.381% |
+| White matter | 648 | 395 | 111 | 732 | 0.174% |
 
 ## Degree distribution
 
 | Metric | In-degree | Out-degree |
 |--------|-----------|------------|
-| Mean | 1.4 | 1.4 |
-| Std | 1.9 | 2.1 |
-| Median | 1.0 | 1.0 |
-| Max | 17 | 24 |
-
-## Network activity
-
-| Metric | Value |
-|--------|-------|
-| Steps | 100 |
-| Mean activation | nan |
-| Max activation | nan |
-| Active neurons | 0/2106 (0%) |
-
-![Simulation](simulation.png)
-![Brain comparison](brain_comparison.png)
-
-
-## Multi-angle views
-
-### Bottom Up
-| Biological | Simulated |
-|---|---|
-| ![bio_bottom_up](views/bio_bottom_up.png) | ![sim_bottom_up](views/sim_bottom_up.png) |
-
-### Front
-| Biological | Simulated |
-|---|---|
-| ![bio_front](views/bio_front.png) | ![sim_front](views/sim_front.png) |
-
-### Rear
-| Biological | Simulated |
-|---|---|
-| ![bio_rear](views/bio_rear.png) | ![sim_rear](views/sim_rear.png) |
-
-### Side
-| Biological | Simulated |
-|---|---|
-| ![bio_side](views/bio_side.png) | ![sim_side](views/sim_side.png) |
-
-### Top
-| Biological | Simulated |
-|---|---|
-| ![bio_top](views/bio_top.png) | ![sim_top](views/sim_top.png) |
-
-
+| Mean | 4.8 | 4.8 |
+| Std | 6.3 | 7.0 |
+| Median | 3.0 | 2.0 |
+| Max | 70 | 124 |
 
 ## Quick start
 
@@ -127,19 +75,28 @@ from safetensors.torch import load_file
 
 # Load everything
 config  = json.load(open("config.json"))
-weights = load_file("connectome.safetensors")["weights"]  # (N, N) synapse counts
+weights = load_file("connectome.safetensors")["weights"]  # (16087, 16087)
 meta    = np.load("metadata.npz", allow_pickle=True)
-edges   = np.load("edges.npz")["edges"]                   # (M, 3) [pre, post, type]
+edges   = np.load("edges.npz")["edges"]                   # (116611, 3)
+
+print(f"{config['n_neurons']} neurons, {config['n_synapses']} connections")
+print(f"Weight matrix: {weights.shape}, density: {config['density']:.4%}")
+```
 
-print(f"{config['n_neurons']} neurons, {config['n_synapses']} synapses")
-print(f"Positions shape: {meta['positions'].shape}")
-print(f"Edge list shape: {edges.shape}")
+### Load via HuggingFace Hub
 
-# Load the initialized network model
-state_dict = load_file("model.safetensors")
+```python
+from huggingface_hub import hf_hub_download
+from safetensors.torch import load_file
+import numpy as np
+
+repo = "NathanRoll/h01-cortex-snn"
+weights = load_file(hf_hub_download(repo, "connectome.safetensors"))["weights"]
+meta = np.load(hf_hub_download(repo, "metadata.npz"), allow_pickle=True)
+print(f"Loaded {weights.shape[0]} neurons")
 ```
 
-### Reconstruct the connectome from the edge list
+### Reconstruct from edge list
 
 ```python
 N = config["n_neurons"]
@@ -149,29 +106,16 @@ for pre, post, stype in edges:
 # W[i, j] = number of synapses from neuron j → neuron i
 ```
 
-### Access neuron metadata
-
-```python
-import pandas as pd
-somas = pd.read_csv("somas_filtered.csv")
-print(somas[["celltype", "layer"]].value_counts())
-```
-
 ## Files
 
-| File | Description | Size hint |
-|------|-------------|-----------|
-| `model.safetensors` | Full model state dict (weights, mask, positions) | ~NxN×4 bytes |
-| `connectome.safetensors` | Raw NxN weight matrix | ~NxN×4 bytes |
-| `edges.npz` | Filtered edge list `[pre, post, type]` | ~M×12 bytes |
-| `metadata.npz` | Positions, types, layers, segment IDs | small |
-| `somas_filtered.csv` | Filtered neuron table from H01 | ~1 MB |
-| `config.json` | All hyperparameters + build stats | small |
+| File | Description | Size |
+|------|-------------|------|
+| `connectome.safetensors` | Full 16,087×16,087 weight matrix | ~1 GB |
+| `edges.npz` | Raw edge list `[pre, post, type]` | ~0.6 MB |
+| `metadata.npz` | Positions, cell types, layers, segment IDs | ~0.3 MB |
+| `somas_filtered.csv` | Neuron table (positions, types, layers) | ~1.1 MB |
+| `config.json` | Build parameters + summary statistics | small |
 | `layer_stats.json` | Per-layer connectivity statistics | small |
-| `camera_params.json` | Exact camera angles for all views | small |
-| `simulation.png` | Activation heatmap + mean activity + traces | ~300 KB |
-| `brain_comparison.png` | Side-by-side bio vs sim (main angle) | ~1 MB |
-| `views/` | Multi-angle 3D renders (bio + sim) | ~5 MB |
 
 ## Data source
 
@@ -186,5 +130,5 @@ microscopy at 4 nm × 4 nm × 33 nm resolution.
 
 ## License
 
-Apache 2.0.  The underlying H01 data is subject to
+Apache 2.0. The underlying H01 data is subject to
 [Google's release terms](https://h01-release.storage.googleapis.com/data.html).

config.json

@@ -23,16 +23,39 @@
     "INTERNEURON"
   ],
   "inhibitory_gain": 4.0,
-  "activation": "relu",
+  "activation": "tanh",
+  "spectral_radius": 0.9,
+  "leak_rate": 0.3,
   "crop_fraction": null,
   "n_steps": 100,
   "input_scale": 0.1,
+  "neuron_model": "rate",
+  "dt_ms": 1.0,
+  "tau_m_exc": 20.0,
+  "tau_m_inh": 10.0,
+  "v_threshold": -50.0,
+  "v_reset": -65.0,
+  "v_rest": -65.0,
+  "tau_w": 150.0,
+  "adaptation_b": 0.08,
+  "experiment_name": "default",
+  "seed": 42,
+  "duration_ms": 5000.0,
+  "warmup_ms": 1000.0,
+  "control_types": [
+    "erdos_renyi",
+    "configuration_model",
+    "spatial_random",
+    "layer_preserving"
+  ],
+  "n_control_instances": 5,
   "device": "cuda",
-  "n_neurons": 2106,
-  "n_synapses": 3041,
-  "n_excitatory": 1466,
-  "n_inhibitory": 475,
-  "density": 0.0006856455890960481,
-  "n_external_inputs": 2576814,
-  "n_edges_raw": 4885
+  "n_neurons": 16087,
+  "n_synapses": 76903,
+  "n_excitatory": 10531,
+  "n_inhibitory": 4688,
+  "density": 0.00029716192183988806,
+  "n_external_inputs": 27022313,
+  "n_edges_raw": 116611,
+  "build": "full_166_shards"
 }

connectome.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d529847f5ed1c356d9865baf2987c10eb3d68a12913442f50cac2d1ecabd21f
-size 17741032
+oid sha256:1df19314f95b35b8bdda3429c8b50c782147f2cc2a32c1182a672c9765ee15fc
+size 1035166364

edges.npz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6207c3ed98524c645a16e593d06b05dab3bb64d1cc1e58bba141d26715455f09
-size 19776
+oid sha256:d685966d1818b32c89f6bd53c0a56a39c4e046899935c9318b250525d2cbd572
+size 627858

layer_stats.json

@@ -2,77 +2,85 @@
   "layers": [
     {
       "layer": "Layer 1",
-      "n_neurons": 48,
-      "n_excitatory": 18,
-      "n_inhibitory": 3,
-      "internal_connections": 0,
-      "density": 0.0
+      "n_neurons": 827,
+      "n_excitatory": 85,
+      "n_inhibitory": 586,
+      "internal_connections": 55,
+      "density": 8e-05
     },
     {
       "layer": "Layer 2",
-      "n_neurons": 120,
-      "n_excitatory": 66,
-      "n_inhibitory": 43,
-      "internal_connections": 51,
-      "density": 0.003542
+      "n_neurons": 4656,
+      "n_excitatory": 2952,
+      "n_inhibitory": 1594,
+      "internal_connections": 21845,
+      "density": 0.001008
     },
     {
       "layer": "Layer 3",
-      "n_neurons": 320,
-      "n_excitatory": 200,
-      "n_inhibitory": 106,
-      "internal_connections": 420,
-      "density": 0.004102
+      "n_neurons": 2692,
+      "n_excitatory": 1673,
+      "n_inhibitory": 965,
+      "internal_connections": 11018,
+      "density": 0.00152
     },
     {
       "layer": "Layer 4",
-      "n_neurons": 1002,
-      "n_excitatory": 748,
-      "n_inhibitory": 196,
-      "internal_connections": 1367,
-      "density": 0.001362
+      "n_neurons": 3440,
+      "n_excitatory": 2622,
+      "n_inhibitory": 688,
+      "internal_connections": 8748,
+      "density": 0.000739
     },
     {
       "layer": "Layer 5",
-      "n_neurons": 441,
-      "n_excitatory": 314,
-      "n_inhibitory": 100,
-      "internal_connections": 385,
-      "density": 0.00198
+      "n_neurons": 2313,
+      "n_excitatory": 1665,
+      "n_inhibitory": 505,
+      "internal_connections": 6252,
+      "density": 0.001169
     },
     {
       "layer": "Layer 6",
-      "n_neurons": 106,
-      "n_excitatory": 80,
-      "n_inhibitory": 13,
-      "internal_connections": 112,
-      "density": 0.009968
+      "n_neurons": 1077,
+      "n_excitatory": 906,
+      "n_inhibitory": 128,
+      "internal_connections": 4419,
+      "density": 0.00381
     },
     {
       "layer": "White matter",
-      "n_neurons": 3,
-      "n_excitatory": 0,
-      "n_inhibitory": 0,
-      "internal_connections": 1,
-      "density": 0.111111
+      "n_neurons": 648,
+      "n_excitatory": 395,
+      "n_inhibitory": 111,
+      "internal_connections": 732,
+      "density": 0.001743
     },
     {
       "layer": "unclassified",
-      "n_neurons": 66,
-      "n_excitatory": 40,
-      "n_inhibitory": 14,
-      "internal_connections": 8,
-      "density": 0.001837
+      "n_neurons": 351,
+      "n_excitatory": 203,
+      "n_inhibitory": 89,
+      "internal_connections": 174,
+      "density": 0.001412
+    },
+    {
+      "layer": "unknown",
+      "n_neurons": 83,
+      "n_excitatory": 30,
+      "n_inhibitory": 22,
+      "internal_connections": 3,
+      "density": 0.000435
     }
   ],
   "degrees": {
-    "in_degree_mean": 1.4,
-    "in_degree_std": 1.9,
-    "in_degree_max": 17,
-    "in_degree_median": 1.0,
-    "out_degree_mean": 1.4,
-    "out_degree_std": 2.1,
-    "out_degree_max": 24,
-    "out_degree_median": 1.0
+    "in_degree_mean": 4.8,
+    "in_degree_std": 6.3,
+    "in_degree_max": 70,
+    "in_degree_median": 3.0,
+    "out_degree_mean": 4.8,
+    "out_degree_std": 7.0,
+    "out_degree_max": 124,
+    "out_degree_median": 2.0
   }
 }

metadata.npz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3bdcb57dda30ee9a15b820e6043c4c032cfd84cb85452624d12de7b288f3ab2c
-size 37330
+oid sha256:ce26bffbb2107bb4c6805d0f52a313d59a6b6b52452807907ef3d9f260e87743
+size 283125