© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

APIAssistant AxisNEW Circuit TracerNEW Steer SAE Evals Exports Community Blog Privacy & Terms Contact

GEMMA-2-2B · 4-CLT-HP · 3248 ｜ Neuronpedia

Home
Gemma-2-2B
4-CLT-HP
3248

INDEX

Explanations

punctuation

np_max-act · gemini-2.0-flash

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

No Configuration Found

Embeds

Show PlotsShow ExplanationShow ActivationsShow Test FieldShow SteerShow Link

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

The

-0.52

ARAN

-0.52

 However

-0.52

We

-0.51

 When

-0.50

 Other

-0.50

ifikant

-0.50

But

-0.48

 There

-0.48

[]:

-0.47

POSITIVE LOGITS

<bos>

1.01

 purpoſe

0.78

 itſelf

0.76

 myſelf

0.71

 raiſ

0.70

 himſelf

0.69

 leaſt

0.66

 uſed

0.66

 whoſe

0.65

 faſt

0.64

Activations Density 1.158%

No Known Activations