-
I'm puzzled by the number of trainable parameters in networks using `gconv2d`.
The script below creates a network using `gconv2d`s from Z2 to C4 to C4 and counts the number of learnable parameters…
-
I really like this E(3)-equivariant transformer. But I have one question about how to achieve SE(3)-equivariant based on your framework. Because I am studying in molecules and proteins where chriality…
-
Two methods to measure equivariance:
- Global: Using [Wang 2022] equivariance error
- Local (layerwise): Using [Gruver 2022] Lie Derivative
Tasks:
- [x] #13
- [x] #18
- [x] #17
- [x] #16
dgcnz updated
4 weeks ago
-
## 🚀 Feature
An implementation of group-equivariant convolutions defined by Cohen et. al in the paper 'Group equivariant convolutions' - in ICML 2016.
Specifically, an implementation of the group-co…
-
Important details:
- script has to checkpoint ~10 times (to be defined)
## Draft description
- [x] get 10 checkpoints for for smokeplume relaxed for the best alpha in wang2022. Use rotation eq…
dgcnz updated
1 month ago
-
In some paper of @gcarleo I saw that, in order to consider symmetries of a Hamiltonian/lattice, one can bring each sample to a canonical form (e.g. lexicographical order of all the equivalent configur…
-
**The bug**
Hi, I see a problem with the code calculating the size of the layers:
In layer_info.py line 109:
```
if hasattr(inputs[0], "size") and callable(inputs[0].size):
return list(i…
-
Evaluations of a symmetric basis with `property=EuclideanMatrix{Float64}`
1. are **symmetric matrice**s as the passing of the following test shows
```julia
@info("Check for symmetry")
for ntest…
-
Hello,
I have a few questions regarding the project. I'm struggling to find how model finds the keypoint values. I found with some code tinkering that you get 10 keypoints as per config that you th…
-
[e3nn](https://www.connectedpapers.com/main/c7215ab4589ce3590910c597e86b1aba8e460d49/e3nn%3A-Euclidean-Neural-Networks/graph)
```mermaid
flowchart TD;
A[Freeman,1991\nThe design and use of stee…