© Neuronpedia 2026

Privacy & Terms Blog GitHub Slack Twitter Contact

Neuronpedia

Jacobian LensNEW

Natural Language

NEW Assistant AxisNEW Circuit TracerUPDATESteer SAE Evals ExportsAPI Community Blog Privacy & Terms Contact

QWEN3-8B · 18-RESID-BATCHTOPK-65K__L0-80 · 1242 ｜ Neuronpedia

Home
Qwen3-8B
18-RESID-BATCHTOPK-65K__L0-80
1242

INDEX

Explanations

study of

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

adamkarvonen/qwen3-8b-saes/saes_Qwen_Qwen3-8B_batch_top_k/resid_post_layer_18

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

ities

-0.30

ArrayOf

-0.24

ality

-0.22

æ¦Ķ

-0.21

itude

-0.19

 ÑģÐ¾Ð±Ð¾Ð¹

-0.19

stay

-0.18

reas

-0.18

ãģªãĤīãģªãģĦ

-0.18

stÃ¤

-0.17

POSITIVE LOGITS

mate

0.25

mates

0.22

tte

0.22

 undertaken

0.21

cape

0.21

hread

0.21

à¸£à¸²à¸§

0.20

äººåĵ¡

0.20

cken

0.20

erge

0.20

Activations Density 0.173%

No Known Activations