INDEX

Explanations

Code "Hello World" examples

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

efd

-0.29

tal

-0.27

enville

-0.24

GENERAL

-0.24

arta

-0.24

esti

-0.23

OfType

-0.23

 spokesman

-0.23

 anticipation

-0.23

Ã©sz

-0.23

POSITIVE LOGITS

Hello

0.78

 Hello

0.75

 hello

0.73

hello

0.68

>Hello

0.62

 HelloWorld

0.58

_hello

0.54

ç¬¬ä¸Ģä¸ª

0.49

Hallo

0.46

_HEL

0.46

Activations Density 0.034%

Code "Hello World" examples

No Comments

No Known Activations