INDEX

Explanations

Computer code/documentation

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_32.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.32.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

icot

-0.27

 Fork

-0.27

ensitive

-0.27

_miss

-0.26

Miss

-0.25

alse

-0.25

compound

-0.25

éĵĥ

-0.25

-cl

-0.25

.operation

-0.25

POSITIVE LOGITS

eme

0.28

ç³»ç»ŁçļĦ

0.28

ç«ł

0.28

 bÃ³

0.27

ling

0.25

 systems

0.25

/init

0.24

ä¿¡æģ¯å®īåħ¨

0.24

æĬł

0.24

Ã¶ne

0.24

Activations Density 0.003%

Computer code/documentation

No Comments

No Known Activations