INDEX

Explanations

a

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_33.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.33.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

åħ»æĪĲ

-0.27

è¶ĬæĿ¥

-0.27

 verm

-0.27

amus

-0.27

/AFP

-0.25

formats

-0.25

schemas

-0.24

åĽ¿

-0.24

 fart

-0.24

 åĪłéĻ¤

-0.24

POSITIVE LOGITS

 Double

0.34

0.31

åıĮ

0.30

 double

0.30

im

0.28

yn

0.28

Double

0.28

}]↵

0.27

 experiment

0.27

Activations Density 0.046%

a

No Known Activations