INDEX

Explanations

Sie/Jewish

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

alf

-0.27

abox

-0.26

issor

-0.25

aec

-0.25

optimize

-0.25

æļĹç¤º

-0.25

met

-0.25

izen

-0.24

select

-0.24

suggest

-0.24

POSITIVE LOGITS

dom

0.28

 contractors

0.24

èĪĨ

0.24

èĻ±

0.24

ç¤ģ

0.24

dom

0.24

 trip

0.24

æºľ

0.23

 trips

0.23

oms

0.23

Activations Density 0.001%

Sie/Jewish

No Comments

No Known Activations