INDEX

Explanations

the

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_27.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.27.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

åħ¶ä¸Ńä¸Ģä¸ª

-0.29

æ³ķåĪĻ

-0.25

åĲĦæĸ¹éĿ¢

-0.25

æ¯ķç«Łæĺ¯

-0.25

 prompts

-0.24

åıĪèĥ½

-0.24

.boost

-0.24

esModule

-0.23

exc

-0.23

 nhiÃªn

-0.23

POSITIVE LOGITS

 allied

0.28

ä¼Łå¤§çļĦ

0.27

 newArray

0.27

quito

0.26

ç´§

0.26

fires

0.26

uits

0.26

entric

0.26

reten

0.26

 Ring

0.25

Activations Density 0.003%

the

No Comments

No Known Activations