INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_22.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.22.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

æŁ¥æĺİ

-0.27

æ±ºå®ļ

-0.26

ypical

-0.26

ðŁĵĸ

-0.25

 Stay

-0.24

ä¾Ŀæĵļ

-0.24

ossier

-0.24

 stay

-0.24

aked

-0.24

åĨ³å®ļäºĨ

-0.24

POSITIVE LOGITS

inc

0.25

.turn

0.24

 substr

0.23

_half

0.23

esis

0.23

 then

0.23

 derog

0.23

_TURN

0.23

 substrate

0.23

hop

0.23

Activations Density 0.002%

No Known Activations

This feature has no known activations.