© 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

QWEN3.5-4B · 15-RES-MATRYOSHKA-65K · 22210 ｜ Neuronpedia

Home
Qwen3.5-4B
15-RES-MATRYOSHKA-65K
22210

INDEX

Explanations

modern adjectives

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

decoderesearch/qwen-3.5-saes/qwen-3.5-4b

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

emblÃ©e

-0.08

guan

-0.08

 geler

-0.08

affichage

-0.08

gressor

-0.07

/Gate

-0.07

 GRAT

-0.07

tees

-0.07

æ²³åĮº

-0.07

æĥļ

-0.07

POSITIVE LOGITS

0.09

an

0.07

0.07

0.07

ashed

0.06

0.06

ized

0.06

0.06

oby

0.06

ashing

0.06

Activations Density 0.009%

No Known Activations