© 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-3-27B · 40-GEMMASCOPE-2-RES-16K · 8838 ｜ Neuronpedia

Home
Gemma-3-27B
40-GEMMASCOPE-2-RES-16K
8838

INDEX

Explanations

solution

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

solutions architect

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

google/gemma-scope-2-27b-pt/resid_post/layer_40_width_16k_l0_medium

Prompts (Dashboard)

392,802 prompts, 256 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

缀

0.52

GEST

0.46

лення

0.44

摫

0.43

ني

0.43

拢

0.43

хран

0.43

राह

0.41

 Casi

0.41

掎

0.41

POSITIVE LOGITS

lø

0.69

RESOL

0.69

 Resolved

0.66

lö

0.62

 resolved

0.61

 RESOL

0.60

 resolución

0.59

 risolvere

0.59

resol

0.58

 разре

0.58

Activations Density 0.069%

No Known Activations