INDEX

Explanations

Roger

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_22.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.22.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ä¸ĭæīĭ

-0.29

stance

-0.28

urus

-0.26

edo

-0.26

ForgeryToken

-0.25

 level

-0.25

sit

-0.25

rap

-0.25

ãĥ¬ãĥ¼ãĤ·ãĥ§ãĥ³

-0.25

åºĬä¸Ĭ

-0.24

POSITIVE LOGITS

ium

0.27

hunt

0.26

bec

0.25

pres

0.24

uen

0.24

åģıå¥½

0.24

hints

0.24

 hunt

0.23

Stan

0.23

(+

0.23

Activations Density 0.348%

Roger

No Comments

No Known Activations