issues
search
chr5tphr
/
zennit
Zennit is a high-level framework in Python using PyTorch for explaining/exploring neural networks using attribution methods like LRP.
Other
183
stars
32
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Docs: Pin nbconvert < 7.14
#207
chr5tphr
closed
3 months ago
0
Core: yield from in collect_leaves
#206
chr5tphr
closed
3 months ago
0
Example: Add option to ignore residual
#205
chr5tphr
closed
3 months ago
0
Potential issue with learning BatchNorm parameters
#204
MikiFER
opened
7 months ago
0
Changing The Value of epsilon in the composite Epsilonplusflat creates a bug
#203
StudentAlessandro
closed
7 months ago
2
Add torchvision.ops.FrozenBatchNorm2d to types
#202
rachtibat
opened
9 months ago
0
SmoothGrad - potential typo in implementation of noise level
#201
annahedstroem
closed
9 months ago
2
Model inference outside of composite context
#200
MikiFER
opened
10 months ago
5
ValueError message in AlphaBeta rule
#199
jackmcrider
closed
10 months ago
1
Neuralized K-Means: make k-means amenable to neural network explanations
#198
jackmcrider
opened
10 months ago
0
Neuralized K-Means
#197
jackmcrider
opened
10 months ago
4
Core: Multiple Inputs, Outputs, and Keyword Arguments
#196
chr5tphr
opened
11 months ago
0
Example: Add ResNet18/VGG11 to feed_forward.py
#195
chr5tphr
closed
10 months ago
0
ResNet: Unstable Attributions
#194
chr5tphr
opened
11 months ago
3
Canonizers: BatchNorm loses bias relevance
#193
chr5tphr
opened
11 months ago
0
Proper way of handling classifier heads in Transformers
#192
kjczarne
opened
11 months ago
2
remove nn.MultiheadAttention from Activation types.py
#191
rachtibat
closed
12 months ago
1
Fix small typo in image-classification-vgg-resnet.ipynb
#190
p16i
closed
12 months ago
1
Binary classification model with one output for output layer
#189
wt12318
closed
1 year ago
1
Bug: MultiheadAttention in types
#188
rachtibat
closed
12 months ago
4
Feature Request: Debugger
#187
rachtibat
opened
1 year ago
2
Added flag to toggle trained weights in tutorial. Also removed unused import. #164
#186
chrissy0
opened
1 year ago
1
Canonizers/ParamMod: Fix Gradients for Parameters
#185
chr5tphr
closed
12 months ago
0
Custom rules for vision transformer
#184
ascdqz
opened
1 year ago
1
Obtanining graident of LRP otput w.r.t. network parameters
#183
MikiFER
closed
12 months ago
12
WIP: GradOutHook, MergedComposite and multi-hooks
#182
chr5tphr
opened
1 year ago
0
Smooth MaxPool2D rule
#181
rachtibat
opened
1 year ago
2
Style: Update PyLint config, add suggestions
#179
chr5tphr
closed
1 year ago
0
Docs: Expand How-To on MixedComposite
#178
chr5tphr
closed
1 year ago
0
Tests: add partial_name_map_composite fixture
#177
chr5tphr
closed
1 year ago
0
Module with Multiple Inputs
#176
rachtibat
opened
1 year ago
2
CI: Fix flake8 update: 3-digit error codes
#175
chr5tphr
closed
1 year ago
0
Docs: Gradient.forward() input not optional
#174
sltzgs
closed
1 year ago
1
Composite: fix passing of beta_smooth value
#173
annahdo
closed
1 year ago
0
Composite: bug in BetaSmooth constructor?
#172
annahdo
closed
1 year ago
3
Densenet canonizations
#171
gumityolcu
opened
1 year ago
0
error occurred when run tutorial on custom model
#170
ascdqz
closed
1 year ago
7
LinearAttention Module
#169
rachtibat
opened
1 year ago
1
Core: Multiple input/param gradient modification
#168
chr5tphr
opened
1 year ago
0
Layer-wise LRP score
#167
shivmgg
opened
1 year ago
6
Docstring/Docs: Fix various typos
#166
chr5tphr
closed
1 year ago
0
How to add support to DenseNet from torchvision
#165
mstaczek
opened
1 year ago
2
Docs: Build Tutorials on ReadTheDocs with trained weights
#164
chr5tphr
opened
1 year ago
0
Example: Add InputXGrad to feed_forward.py
#163
chr5tphr
closed
1 year ago
0
Docstring: Canonizer: Fix wrong reST
#162
chr5tphr
closed
1 year ago
0
Tests: Fix SmoothGrad variance check
#161
chr5tphr
closed
1 year ago
0
MixedComposite and NameLayerMapComposite for combining arbitrary Composites
#160
dkrako
closed
1 year ago
9
Core: Second Order Gradients
#159
chr5tphr
closed
1 year ago
9
Fix `f-string` in an errror message
#158
p16i
closed
1 year ago
1
Core/Composite: Simplify Composite.context
#156
chr5tphr
opened
1 year ago
0
Next