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

ovsky

-0.32

 hues

-0.28

åĨįè¯´

-0.27

 alternate

-0.27

urus

-0.26

 whom

-0.26

ÇĲ

-0.26

Alternate

-0.25

own

-0.24

 ÑĤÐ°Ðº

-0.24

POSITIVE LOGITS

umps

0.26

éģįåıĬ

0.26

rosso

0.26

 Stard

0.25

 strokeLine

0.25

ä¹ĭå¿ĥ

0.25

 squeez

0.25

ections

0.24

 Claw

0.24

.qq

0.24

Activations Density 0.177%

No Known Activations

This feature has no known activations.