INDEX

Explanations

the

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_30.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.30.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

æĸ°éĹ»è®°èĢħ

-0.32

-0.31

 presently

-0.30

&o

-0.30

Ã¦

-0.30

tho

-0.29

chÃ¨

-0.28

æħ¢æħ¢çļĦ

-0.28

è±«

-0.28

))^

-0.27

POSITIVE LOGITS

åĺĽ

0.32

åĪļå¼Ģå§ĭ

0.30

æİ¥ä¸ĭæĿ¥

0.28

åĳ¨è¾¹

0.28

 basically

0.28

ç»Ħå»º

0.28

 incremental

0.27

çĻ¾åĪĨçĻ¾

0.26

 unstoppable

0.26

ï¼Ĺ

0.26

Activations Density 0.889%

the

No Known Activations