Spaces:
Runtime error
Runtime error
Alexander Seifert
commited on
Commit
•
408486e
1
Parent(s):
4c13f39
improve docs
Browse files- README.md +1 -1
- html/index.md +1 -1
- src/subpages/hidden_states.py +8 -8
README.md
CHANGED
@@ -24,7 +24,7 @@ A group of neurons tend to fire in response to commas and other punctuation. Oth
|
|
24 |
|
25 |
### Embeddings
|
26 |
|
27 |
-
For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with
|
28 |
|
29 |
|
30 |
### Probing
|
|
|
24 |
|
25 |
### Embeddings
|
26 |
|
27 |
+
For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with disagreements marked by a small black border.
|
28 |
|
29 |
|
30 |
### Probing
|
html/index.md
CHANGED
@@ -51,7 +51,7 @@ Activations
|
|
51 |
|
52 |
Hidden States
|
53 |
|
54 |
-
> For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with
|
55 |
>
|
56 |
> Using these projections you can visually identify data points that end up in the wrong neighborhood, indicating prediction/labeling errors.
|
57 |
|
|
|
51 |
|
52 |
Hidden States
|
53 |
|
54 |
+
> For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with disagreements marked by a small black border.
|
55 |
>
|
56 |
> Using these projections you can visually identify data points that end up in the wrong neighborhood, indicating prediction/labeling errors.
|
57 |
|
src/subpages/hidden_states.py
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
"""
|
2 |
-
For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with
|
3 |
"""
|
4 |
import numpy as np
|
5 |
import plotly.express as px
|
@@ -86,7 +86,7 @@ class HiddenStatesPage(Page):
|
|
86 |
|
87 |
with st.expander("💡", expanded=True):
|
88 |
st.write(
|
89 |
-
"For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with
|
90 |
)
|
91 |
|
92 |
col1, _, col2 = st.columns([9 / 32, 1 / 32, 22 / 32])
|
@@ -152,12 +152,12 @@ class HiddenStatesPage(Page):
|
|
152 |
df["sent2"] = df["ids"].map(lambda x: " ".join(sents[x][100:150].split()))
|
153 |
df["sent3"] = df["ids"].map(lambda x: " ".join(sents[x][150:200].split()))
|
154 |
df["sent4"] = df["ids"].map(lambda x: " ".join(sents[x][200:250].split()))
|
155 |
-
df["
|
156 |
|
157 |
subset = df[:n_tokens]
|
158 |
-
|
159 |
-
x=subset[subset["
|
160 |
-
y=subset[subset["
|
161 |
mode="markers",
|
162 |
marker=dict(
|
163 |
size=6,
|
@@ -178,7 +178,7 @@ class HiddenStatesPage(Page):
|
|
178 |
hover_name="tokens",
|
179 |
title="Colored by label",
|
180 |
)
|
181 |
-
fig.add_trace(
|
182 |
st.plotly_chart(fig)
|
183 |
|
184 |
fig = px.scatter(
|
@@ -190,5 +190,5 @@ class HiddenStatesPage(Page):
|
|
190 |
hover_name="tokens",
|
191 |
title="Colored by prediction",
|
192 |
)
|
193 |
-
fig.add_trace(
|
194 |
st.plotly_chart(fig)
|
|
|
1 |
"""
|
2 |
+
For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with disagreements marked by a small black border.
|
3 |
"""
|
4 |
import numpy as np
|
5 |
import plotly.express as px
|
|
|
86 |
|
87 |
with st.expander("💡", expanded=True):
|
88 |
st.write(
|
89 |
+
"For every token in the dataset, we take its hidden state and project it onto a two-dimensional plane. Data points are colored by label/prediction, with disagreements signified by a small black border."
|
90 |
)
|
91 |
|
92 |
col1, _, col2 = st.columns([9 / 32, 1 / 32, 22 / 32])
|
|
|
152 |
df["sent2"] = df["ids"].map(lambda x: " ".join(sents[x][100:150].split()))
|
153 |
df["sent3"] = df["ids"].map(lambda x: " ".join(sents[x][150:200].split()))
|
154 |
df["sent4"] = df["ids"].map(lambda x: " ".join(sents[x][200:250].split()))
|
155 |
+
df["disagreements"] = df["labels"] != df["preds"]
|
156 |
|
157 |
subset = df[:n_tokens]
|
158 |
+
disagreements_trace = go.Scatter(
|
159 |
+
x=subset[subset["disagreements"]]["x"],
|
160 |
+
y=subset[subset["disagreements"]]["y"],
|
161 |
mode="markers",
|
162 |
marker=dict(
|
163 |
size=6,
|
|
|
178 |
hover_name="tokens",
|
179 |
title="Colored by label",
|
180 |
)
|
181 |
+
fig.add_trace(disagreements_trace)
|
182 |
st.plotly_chart(fig)
|
183 |
|
184 |
fig = px.scatter(
|
|
|
190 |
hover_name="tokens",
|
191 |
title="Colored by prediction",
|
192 |
)
|
193 |
+
fig.add_trace(disagreements_trace)
|
194 |
st.plotly_chart(fig)
|