INDEX

Explanations

code and statistics

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

Not in Any Lists

No Comments

Negative Logits

abol

-0.28

 Bett

-0.27

 people

-0.27

ç¾¤ä½ĵ

-0.26

<?↵

-0.26

bury

-0.26

 Brothers

-0.25

äºº

-0.24

 group

-0.24

 Boys

-0.24

POSITIVE LOGITS

ãĥªãĥ³ãĤ°

0.27

venir

0.27

ä¸´æ¸¯

0.26

èĭ¥æĤ¨

0.25

IClient

0.25

zent

0.25

çĿ¹

0.24

-guard

0.24

ä¹Łä¸įä¾ĭå¤ĸ

0.24

ÑıÐµÑĤ

0.24

Activations Density 0.001%

code and statistics

No Comments

No Known Activations