N Neurarch Architectures Models Checks Data Docs Open the app

Comparisons / Neural Collaborative Filtering vs Neural Collaborative Filtering

Neural Collaborative Filtering vs Neural Collaborative Filtering

Neural collaborative filtering against its generalised-matrix-factorisation hybrid.

Neural Collaborative Filtering has 70M more parameters than Neural Collaborative Filtering: 4 layers added, 5 changed.

Baseline

Neural Collaborative Filtering

Layers
9
Parameters
35M
Input
1
Output
1 × 1
Forward-passes
yes
Est. train cost
$0.053
Compared

Neural Collaborative Filtering

Layers
13
Parameters
106M
Input
1
Output
1 × 1
Forward-passes
yes
Est. train cost
$0.076

The deltas

Every number is Neural Collaborative Filtering relative to Neural Collaborative Filtering.

Parameters
+70M (+200%)
Layers
+4
Added
4
Removed
0
Changed
5
Unchanged
7

Which GPUs each one fits

Memory for the graph at the input shape both declare. A highlighted row is a card one of them fits and the other does not, which is the difference that decides a purchase.

GPUNeural Collaborative FilteringNeural Collaborative Filtering
T4 16GBfitsfits
A100 40GBfitsfits
H100 80GBfitsfits

Layer by layer

Aligned in topological order. 7 of 16 rows are the same layer with the same parameters.

Hide all 16 rows
Neural Collaborative FilteringNeural Collaborative Filtering
LayerParamsOutputLayerParamsOutput
1sameUser ID
Input
1User ID
Input
1
2sameItem ID
Input
1Item ID
Input
1
3addedUser MF Emb
Embedding
3.2M1 × 32
4addedUser MLP Emb
Embedding
6.4M1 × 64
5changed
vocabSize
User Embedding
Embedding
3.2M1 × 32Item MF Emb
Embedding
32M1 × 32
6changed
embeddingDim
Item Embedding
Embedding
32M1 × 32Item MLP Emb
Embedding
64M1 × 64
7addedGMF (⊙)
Multiply
1 × 32
8changed
dim, numInputs, axis
Concat [u; i]
Concatenate
1 × 64Concat
Concatenate
1 × 128
9changed
inFeatures
MLP 1
Linear
4.2K1 × 64MLP FC 1
Linear
8.3K1 × 64
10–12same3 unchanged layers
13addedFuse GMF+MLP
Concatenate
1 × 64
14changed
inFeatures
Score Head
Linear
331 × 1Predict
Linear
651 × 1
15sameSigmoid
Sigmoid
1 × 1Sigmoid
Sigmoid
1 × 1
16sameP(click)
Output
1 × 1P(rating)
Output
1 × 1

What this is not

Take it further

Open either graph in the editor, change it, and check it again: Neural Collaborative Filtering · Neural Collaborative Filtering

Compare any two models of your own, including anything on Hugging Face: the comparison tool.

Machine-readable: this page as markdown · the pair index · POST https://www.neurarch.com/api/v1/plan for a graph of your own.

Related comparisons

Wide & Deep vs Neural Collaborative Filtering
The memorisation-plus-generalisation recommender against a purely neural one.
Two-Tower vs Neural Collaborative Filtering
Retrieval by dot product against a learned interaction.