INDEX

Explanations

Scraped internet data

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_23.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.23.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æĹ¢

-0.31

')."

-0.29

æ¯Ķ

-0.28

åĶ¯

-0.26

åĨį

-0.25

åħĪ

-0.25

#End

-0.25

æĹ¢æĺ¯

-0.24

ocity

-0.24

 unchecked

-0.24

POSITIVE LOGITS

ãĢģãĢģ

0.44

ãĢģ

0.38

ï½¤

0.33

,,,,,,,,

0.29

,,,,

0.29

CD

0.26

ãĢģ“

0.24

ä¼ģä¸ļåĴĮ

0.24

æł²

0.24

Cd

0.24

Activations Density 0.002%

Scraped internet data

No Comments

No Known Activations