INDEX

Explanations

ep

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_31.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.31.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 talking

-0.29

ten

-0.27

æ±ī

-0.26

%=

-0.26

 talks

-0.26

unar

-0.26

ç¢³

-0.25

èµ¶

-0.25

è°Īè¯Ŀ

-0.25

çĻ»ä¸Ĭ

-0.25

POSITIVE LOGITS

iph

0.38

IPH

0.31

è£Ĥ

0.29

imeType

0.27

 Myth

0.26

_MI

0.26

 sÃ¦

0.26

eld

0.26

ate

0.25

sut

0.25

Activations Density 0.003%

ep

No Comments

No Known Activations