© 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

QWEN3-32B · 32-RESID-BATCHTOPK-65K · 20322 ｜ Neuronpedia

Home
Qwen3-32B
32-RESID-BATCHTOPK-65K
20322

INDEX

Explanations

quantities of types

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

èº«

-0.10

ifier

-0.09

 Barnett

-0.09

æĶ¶èİ·

-0.08

McB

-0.08

 Hutch

-0.08

HC

-0.08

.truth

-0.08

.business

-0.08

gett

-0.08

POSITIVE LOGITS

ç±»åŀĭçļĦ

0.12

ç±»åŀĭ

0.11

 types

0.11

types

0.10

 kinds

0.09

ç§įç±»

0.08

 Types

0.08

 ç±»

0.08

ä¸¤ç±»

0.08

éĻĲãĤĬ

0.08

Activations Density 0.591%

No Known Activations