© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

Jacobian LensNEW

Natural Language

NEW Assistant AxisNEW Circuit TracerUPDATESteer SAE Evals Exports interp-engineNEWAPI Community Blog Privacy & Terms Contact

GEMMA-2-27B · 34-GEMMASCOPE-RES-131K · 92700 ｜ Neuronpedia

Home
Gemma-2-27B
34-GEMMASCOPE-RES-131K
92700

INDEX

Explanations

various languages

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Configuration

google/gemma-scope-27b-pt-res/layer_34/width_131k

Prompts (Dashboard)

24,576 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

 sooooo

-1.13

!!

-1.12

 valsty

-1.09

 While

-0.99

 haphazard

-0.98

!!!

-0.97

 frightening

-0.96

 ONLY

-0.95

all

-0.95

or

-0.94

POSITIVE LOGITS

《

1.21

奩

1.12

canActivate

1.11

《

1.11

during

1.11

NFL

1.10

 ultime

1.09

ฟ

1.08

Nigeria

1.07

É

1.07

Activations Density 0.004%

No Known Activations