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
Joseph Bloom · Initial Gemma-2B SAEs
Gemma-2B
Residual Stream
0-RES-JB
230
Prev
Next
MODEL
0-res-jb
Source/SAE
INDEX
Go
Explanations
phrases related to computer code or instructions
oai_token-act-pair · gpt-3.5-turbo
Triggered by @bot
No Scores
New Auto-Interp
AutoInterp Type
claude-4-5-haiku
Generate
Top Features by Cosine Similarity
Configuration
jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic
How To Load
Prompts (Dashboard)
24,576 prompts, 128 tokens each
Dataset (Dashboard)
HuggingFaceFW/fineweb
Features
16,384
Data Type
torch.float32
Hook Point
blocks.0.hook_resid_post
Architecture
standard
Context Size
1,024
Dataset
HuggingFaceFW/fineweb
Hook Point Layer
0
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/gemma-2b/0-res-jb/230?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/gemma-2b/0-res-jb/230?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
50
+0.12
0.5%
90
+0.06
0.3%
1296
+0.06
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1051
+0.12
0.04
1470
+0.06
0.04
1582
+0.06
0.04
Negative Logits
<bos>
-1.61
ⓧ
-0.98
-0.85
<?
-0.82
/**
-0.80
/***
-0.78
/*
-0.69
/*++
-0.67
<?
-0.61
expel
-0.61
POSITIVE LOGITS
code
2.02
Code
1.88
code
1.76
codes
1.76
Code
1.75
CODE
1.73
Codes
1.63
CODE
1.57
codes
1.48
Codes
1.46
Act
ivations
Density 0.116%
Stacked
Snippet
Full
Show Raw Tokens
Show Formatted
Show Breaks
Hide Breaks
No Known Activations
