© 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

QWEN3-32B · 32-RESID-BATCHTOPK-65K · 18704 ｜ Neuronpedia

Home
Qwen3-32B
32-RESID-BATCHTOPK-65K
18704

INDEX

Explanations

shelter

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32

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

-0.11

keit

-0.10

åĦ¿

-0.09

ðŁı»

-0.09

eries

-0.09

ex

-0.09

êº¼

-0.09

ities

-0.09

imizer

-0.09

uset

-0.09

POSITIVE LOGITS

belt

0.10

acked

0.10

ÑĥÑħ

0.10

çĦ¶æĺ¯

0.09

mates

0.09

edata

0.09

acular

0.09

ä¸įä½ı

0.09

edReader

0.09

à¸ªà¸²à¸ĺà¸²à¸£

0.09

Activations Density 0.006%

No Known Activations