© 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-4-31B · 30-RES-MATRYOSHKA-131K · 116819 ｜ Neuronpedia

Home
Gemma-4-31B
30-RES-MATRYOSHKA-131K
116819

INDEX

Explanations

return statements

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

decoderesearch/gemma-4-saes/gemma-4-31b

Prompts (Dashboard)

16,384 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

好事

-0.06

 Good

-0.06

 uitle

-0.06

cribable

-0.06

 Putt

-0.06

 estuary

-0.05

 Dessa

-0.05

 نحاول

-0.05

 enhancing

-0.05

Good

-0.05

POSITIVE LOGITS

创造

0.07

跟我

0.06

 বিকাল

0.06

創造

0.06

 Already

0.06

Already

0.06

ampaikan

0.06

सम्म

0.05

问道

0.05

 সৈন্যরা

0.05

Activations Density 0.001%

No Known Activations