© 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 · 31-GEMMASCOPE-2-RES-262K · 15582 ｜ Neuronpedia

Home
Gemma-3-27B
31-GEMMASCOPE-2-RES-262K
15582

INDEX

Explanations

kill or slay

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

act of killing

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_31_width_262k_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

ס

0.63

 বিষয়ে

0.63

 বিষয়ে

0.59

 undergone

0.58

opération

0.57

 adı

0.55

attiyam

0.55

 ساين

0.55

ഞ്ച്

0.55

র্যের

0.54

POSITIVE LOGITS

 killing

1.02

 kill

1.02

殺

0.99

killing

0.91

 KILL

0.91

 Killing

0.86

 giết

0.85

 killings

0.84

杀

0.84

 killed

0.83

Activations Density 0.047%

No Known Activations