© 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 · 10-GEMMASCOPE-RES-131K · 112954 ｜ Neuronpedia

Home
Gemma-2-27B
10-GEMMASCOPE-RES-131K
112954

INDEX

Explanations

startup business

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_10/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

 FÜR

-1.91

 ות

-1.84

 OGSÅ

-1.83

苙

-1.82

↵

-1.73

 ligue

-1.70

 mangá

-1.66

PÅ

-1.65

 haja

-1.55

 формате

-1.55

POSITIVE LOGITS

 tohoto

1.64

?”

1.64

</i>

1.60

 апреля

1.56

 数据库

1.55

”

1.53

 AppCompatTheme

1.52

 基础

1.50

Sabato

1.50

 结构

1.47

Activations Density 0.012%

No Known Activations