INDEX

Explanations

Health

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

é¼Ĳ

-0.28

duct

-0.25

æ¨¡æĭŁ

-0.24

atism

-0.24

(acc

-0.24

ician

-0.24

æĿĳå§Ķä¼ļ

-0.24

dal

-0.24

åħ¬åı¸çļĦ

-0.24

olean

-0.23

POSITIVE LOGITS

StringEncoding

0.25

çĹķ

0.24

 Notices

0.24

ectors

0.23

isk

0.23

EEK

0.23

 Equal

0.23

 Skip

0.23

 Loose

0.23

lovak

0.22

Activations Density 0.001%

Health

No Comments

No Known Activations