Neuronpedia
Get Started
API
Releases
Jump To
Search
Models
Assistant Axis
NEW
Circuit Tracer
NEW
Steer
SAE Evals
Exports
Community
Blog
Privacy & Terms
Contact
Sign In
© Neuronpedia 2025
Privacy & Terms
Blog
GitHub
Slack
Twitter
Contact
Home
Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
Pythia-70M-Deduped
MLP Post
0-MLP-SM
27767
Prev
Next
MODEL
0-mlp-sm
Source/SAE
INDEX
Go
Explanations
specific numeric values, particularly those related to counts or indices
oai_token-act-pair · gpt-4o-mini
Triggered by @bot
No Scores
New Auto-Interp
AutoInterp Type
claude-4-5-haiku
Generate
Top Features by Cosine Similarity
Configuration
ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm
How To Load
Prompts (Dashboard)
32,768 prompts, 128 tokens each
Dataset (Dashboard)
monology/pile-uncopyrighted
Features
32,768
Data Type
torch.float32
Hook Name
blocks.0.hook_mlp_out
Hook Layer
0
Architecture
standard
Context Size
128
Dataset
EleutherAI/the_pile_deduplicated
Activation Function
relu
Show All
Embeds
Show Plots
Show Explanation
Show Activations
Show Test Field
Show Steer
Show Link
IFrame
<iframe src="https://www.neuronpedia.org/pythia-70m-deduped/0-mlp-sm/27767?embed=true&embedexplanation=true&embedplots=true&embedsteer=true&embedactivations=true&embedlink=true&embedtest=true" title="Neuronpedia" style="height: 300px; width: 540px;"></iframe>
Link
https://www.neuronpedia.org/pythia-70m-deduped/0-mlp-sm/27767?embed=true&embedexplanation=true&embedplots=true&embedsteer=true&embedactivations=true&embedlink=true&embedtest=true
Not in Any Lists
Add to List
▼
No Comments
ADD
Neuron Alignment
Index
Value
% of L₁
307
+0.15
0.9%
369
+0.13
0.7%
6
+0.12
0.7%
Correlated Neurons
Index
P. Corr.
Cos Sim.
307
+0.15
0.04
445
+0.13
0.03
199
+0.12
0.03
Negative Logits
whom
-1.71
king
-1.59
eligible
-1.56
angers
-1.52
chers
-1.51
dition
-1.46
sheaf
-1.45
rib
-1.45
yles
-1.43
aned
-1.36
POSITIVE LOGITS
ª
2.59
®
2.35
Į
2.35
¬
2.26
¹
2.21
¨
2.16
ĨĴ
2.07
¶
1.95
¦
1.87
Ĺ
1.80
Act
ivations
Density 0.150%
Stacked
Snippet
Full
Show Raw Tokens
Show Formatted
Show Breaks
Hide Breaks
No Known Activations
