issues
search
saprmarks
/
dictionary_learning
MIT License
142
stars
37
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
One-dimensional Signal
#27
zyy200196
opened
2 weeks ago
0
Fix BatchTopKSAE training
#26
mntss
opened
4 weeks ago
0
How to train AutoencoderNew?
#25
antklen
opened
1 month ago
0
Missing License
#24
kenza-bouzid
closed
2 weeks ago
1
Log trained SAEs to wandb artifacts
#23
pleask
opened
1 month ago
0
Implement jumprelu training
#22
pleask
closed
2 months ago
2
Use separate wandb runs for each SAE being trained
#21
pleask
closed
2 months ago
1
Implement BatchTopK
#20
pleask
closed
2 months ago
0
JumpReLU SAE training on the roadmap?
#19
ejmichaud
opened
3 months ago
2
small bug fixes
#18
ElanaPearl
closed
3 months ago
1
Merge Collab Branch
#17
canrager
closed
4 months ago
0
Add option to train Gated SAEs
#16
koayon
opened
6 months ago
1
fixed p calculation
#15
canrager
closed
6 months ago
0
added annealing and trainer_param_callback
#14
canrager
closed
6 months ago
0
add ListaTrainer
#13
jannik-brinkmann
closed
6 months ago
0
added sparsity trainer with quadratic activation and p-norm
#12
rangell
closed
4 months ago
0
Fix tutorial 404
#11
adamkarvonen
closed
4 months ago
0
Add sae_set_name to local_path for dictionary downloader
#10
adamkarvonen
closed
7 months ago
0
Tensors must have same number of dimensions: got 2 and 3
#9
joelburget
closed
7 months ago
1
`attention_mask` is required
#8
joelburget
closed
7 months ago
3
`ActivationBuffer` used inconsistently re running out of data.
#7
joelburget
closed
8 months ago
1
add ability to vary more hyperparamaters in parallel training
#6
g-w1
opened
8 months ago
0
Update to use nnsight 0.2 syntax
#5
JadenFiotto-Kaufman
closed
7 months ago
0
NNsight v0.2 Updates
#4
cadentj
closed
8 months ago
0
add an implementation of ghost gradients
#3
g-w1
closed
9 months ago
1
`EmptyStream` isn't used
#2
joelburget
closed
10 months ago
1
Update README.md
#1
cadentj
closed
11 months ago
0