© 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 · 78710 ｜ Neuronpedia

Home
Gemma-2-27B
34-GEMMASCOPE-RES-131K
78710

INDEX

Explanations

template parts

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

 trovano

-0.94

 avrebbe

-0.93

 照片

-0.92

起来

-0.90

はいけない

-0.89

-0.88

 такъ

-0.88

hamento

-0.87

 blunder

-0.86

erweise

-0.86

POSITIVE LOGITS

Tem

1.33

TEM

1.26

tempted

1.25

Tem

1.23

TEM

1.00

 Template

0.98

 tempi

0.97

 documento

0.95

 می‌کند

0.94

atem

0.93

Activations Density 0.015%

No Known Activations