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

ÅĲ

-0.26

ä»»ä½ķæĹ¶åĢĻ

-0.26

ÑĢÐ¼Ð°

-0.25

éķ¿ä¹ħ

-0.25

xce

-0.24

eacher

-0.24

 Downs

-0.24

éĻĶ

-0.24

 ache

-0.23

åıĳèµ·

-0.23

POSITIVE LOGITS

har

0.30

olicies

0.27

direction

0.26

éĿ¢ä¸Ĭ

0.26

ass

0.26

']],

0.25

æ¸´

0.25

time

0.24

 Removal

0.24

ston

0.24

Activations Density 0.030%

No Known Activations

This feature has no known activations.