Architectures / Recommendation
💡 LightGCN
He et al. 2020 — user/item embeddings propagated through 3 light graph-conv layers, layer combination via mean (no transforms, no nonlinearities)
From He et al. (2020). LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation. SIGIR 2020. This page is the graph, not the PDF: open it, edit it, verify it, export it.
Every number on this page is computed from the graph by the same functions the app runs, not written by hand.
When to pick it
Structure
14 layers. Output shapes are propagated from the input shape, batch dimension excluded.
| Layer | Type | Parameters | Output shape | |
|---|---|---|---|---|
| 1 | User ID | Input | shape=[1] | 1 |
| 2 | E_user(0) | Embedding | vocabSize=100000 | 1 × 64 |
| 3 | GraphConv(1) | GraphConv | outFeatures=64, inFeatures=64 | 1 × 64 |
| 4 | GraphConv(2) | GraphConv | outFeatures=64, inFeatures=64 | 1 × 64 |
| 5 | GraphConv(3) | GraphConv | outFeatures=64, inFeatures=64 | 1 × 64 |
| 6 | Layer Combine (mean) | Mean | 64 | |
| 7 | Item ID | Input | shape=[1] | 1 |
| 8 | E_item(0) | Embedding | vocabSize=1000000 | 1 × 64 |
| 9 | GraphConv(1) | GraphConv | outFeatures=64, inFeatures=64 | 1 × 64 |
| 10 | GraphConv(2) | GraphConv | outFeatures=64, inFeatures=64 | 1 × 64 |
| 11 | GraphConv(3) | GraphConv | outFeatures=64, inFeatures=64 | 1 × 64 |
| 12 | Layer Combine (mean) | Mean | 64 | |
| 13 | Dot Score | MatMul | 64 | |
| 14 | Score | Output | 64 |
What the verifier says
The same 43 single-graph checks that run on every edit in the app, on this graph.
deep-no-norm
The PyTorch it exports
Generated from the graph above. First 46 lines; the app exports the whole file, plus the training loop, the data contract and a deploy bundle.
# Architecture designed with Neurarch: https://neurarch.com
# PyTorch: compatible with Python 3.8+ and torch>=1.12
# Colab: pip install torch torchvision (usually pre-installed)
import torch
import torch.nn as nn
import torch.nn.functional as F
from typing import Tuple
from torch_geometric.nn import GCNConv, GATConv, SAGEConv
class LightGCN(nn.Module):
def __init__(self):
super().__init__()
self.embedding_1 = nn.Embedding(100000, 64)
self.graphConv_1 = GCNConv(64, 64)
self.graphConv_2 = GCNConv(64, 64)
self.graphConv_3 = GCNConv(64, 64)
self.embedding_2 = nn.Embedding(1000000, 64)
self.graphConv_4 = GCNConv(64, 64)
self.graphConv_5 = GCNConv(64, 64)
self.graphConv_6 = GCNConv(64, 64)
def forward(self, src, tgt=None, edge_index=None):
# User ID shape: [1]
# Item ID shape: [1]
embedding_ser_e0 = self.embedding_1(src)
graph_conv_er_gc1 = self.graphConv_1(embedding_ser_e0, edge_index)
graph_conv_er_gc2 = self.graphConv_2(graph_conv_er_gc1, edge_index)
graph_conv_er_gc3 = self.graphConv_3(graph_conv_er_gc2, edge_index)
mean_r_mean = torch.stack([embedding_ser_e0, graph_conv_er_gc1, graph_conv_er_gc2, graph_conv_er_gc3]).mean(dim=0)
embedding_tem_e0 = self.embedding_2(tgt)
graph_conv_em_gc1 = self.graphConv_4(embedding_tem_e0, edge_index)
graph_conv_em_gc2 = self.graphConv_5(graph_conv_em_gc1, edge_index)
graph_conv_em_gc3 = self.graphConv_6(graph_conv_em_gc2, edge_index)
mean_m_mean = torch.stack([embedding_tem_e0, graph_conv_em_gc1, graph_conv_em_gc2, graph_conv_em_gc3]).mean(dim=0)
matmul_dot = (mean_r_mean * mean_m_mean).sum(dim=-1, keepdim=True)
# Output
return matmul_dot
if __name__ == '__main__':
model = LightGCN()
model.eval()
edge_index = torch.tensor([[0, 1, 2, 3], [1, 2, 3, 0]], dtype=torch.long) # (2, num_edges) COOFor agents
This architecture is machine-readable end to end. An agent can list the set, fetch this graph, edit it, and have the edit verified before any GPU time is spent.