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

-alpha

-0.27

å¸Ĥ

-0.26

ursor

-0.26

))*

-0.25

å¹´åºķ

-0.25

@m

-0.25

ÙĦØ§ÙĨ

-0.24

orne

-0.24

éĻĦè¿ĳ

-0.24

åľº

-0.24

POSITIVE LOGITS

 Straight

0.30

 geomet

0.29

Dims

0.28

 Madness

0.28

Ill

0.27

 Quad

0.25

 quadr

0.25

rets

0.25

Boh

0.24

æīĭæĦŁ

0.24

Activations Density 0.001%

No Known Activations

This feature has no known activations.