INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_23.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.23.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

leave

-0.29

ãĥ¡ãĥĭãĥ¥ãĥ¼

-0.27

vd

-0.27

.transition

-0.26

otti

-0.26

è¿ĩæ¸¡

-0.26

 subs

-0.25

odon

-0.25

 Leave

-0.25

æĻ¨æĬ¥

-0.24

POSITIVE LOGITS

 mechanically

0.25

ç»Ļåĩº

0.25

Sortable

0.25

.easy

0.25

è§Ĩè§Ĵ

0.24

 somehow

0.24

åĪ»

0.24

ibly

0.24

ä¹ĭåĪĿ

0.24

 sortable

0.24

Activations Density 0.030%

No Known Activations

This feature has no known activations.