Skip to content

fix(visualize): resolve crashes and blank L1 weight plots #320

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
barnaby-wreford wants to merge 2 commits into
base: master
Choose a base branch
from
Open

fix(visualize): resolve crashes and blank L1 weight plots #320

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e8da0ff
fix(visualize): correct num_ls_buckets access and indexing
barnaby-wreford Jun 23, 2025
c02c6dc
fix(visualize): truncate l1 weights to avoid blank plots
barnaby-wreford Jun 23, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 7 additions & 7 deletions visualize.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -317,7 +317,7 @@ def format_coord(x, y):

def plot_fc_weights(self):
if not self.args.no_fc_weights:
num_buckets = self.model.feature_set.num_ls_buckets
num_buckets = self.model.num_ls_buckets
fig, axs = plt.subplots(3, num_buckets, dpi=self.dpi)

extra_info = ""
Expand All @@ -343,9 +343,7 @@ def get_l1_weights(bucket_id, l1):

for i in range(N):
l1_weights[2 * i] = l1_weights_[i][self.sorted_input_neurons]
l1_weights[2 * i + 1] = l1_weights_[i][
self.M + self.sorted_input_neurons
]
l1_weights[2 * i + 1] = l1_weights_[i+N][self.sorted_input_neurons]
return l1_weights, N

def get_l2_weights(bucket_id, l2):
Expand Down Expand Up @@ -376,6 +374,8 @@ def get_l2_weights(bucket_id, l2):
self.model.layer_stacks.get_coalesced_layer_stacks()
):
l1_weights, N = get_l1_weights(bucket_id, l1)
# truncate to prevent matshow rendering a blank plot
truncated_l1_weights = l1_weights[:, :16]
l2_weights = get_l2_weights(bucket_id, l2)
output_weights = output.weight.data.numpy()

Expand All @@ -384,7 +384,7 @@ def get_l2_weights(bucket_id, l2):

ax = axs[0, bucket_id]
im = ax.matshow(
np.abs(l1_weights) if plot_abs else l1_weights,
np.abs(truncated_l1_weights) if plot_abs else truncated_l1_weights,
vmin=vmin,
vmax=vmax,
cmap=cmap,
Expand All @@ -410,7 +410,7 @@ def get_l2_weights(bucket_id, l2):
)

row_names = ["bucket {}".format(i) for i in range(num_buckets)]
col_names = ["l1", "l2", "output"]
col_names = ["l1 (truncated)", "l2", "output"]
for i in range(3):
for j in range(num_buckets):
ax = axs[i, j]
Expand Down Expand Up @@ -482,7 +482,7 @@ def plot_fc_biases(self):
self.ref_model.layer_stacks.get_coalesced_layer_stacks()
)

num_buckets = self.model.feature_set.num_ls_buckets
num_buckets = self.model.num_ls_buckets
fig, axs = plt.subplots(3, num_buckets, dpi=self.dpi)
extra_info = ""
if self.args.sort_input_neurons:
Expand Down