© 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 · 58485 ｜ Neuronpedia

Home
Gemma-2-27B
10-GEMMASCOPE-RES-131K
58485

INDEX

Explanations

expiration notices

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

瑧

-3.44

т

-3.20

㈨

-2.75

 baumwolle

-2.72

ܜ

-2.72

诒

-2.69

 nahil

-2.67

苳

-2.66

注意事

-2.64

憍

-2.61

POSITIVE LOGITS

 Instead

3.06

<bos>

3.02

2.89

 Rather

2.81

 Similarly

2.78

 They

2.75

 Therefore

2.73

 Thus

2.67

 Meanwhile

2.64

2.59

Activations Density 0.002%

No Known Activations