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

ç§¯ç´¯

-0.27

orsch

-0.26

ç´¯ç§¯

-0.26

pac

-0.25

(suffix

-0.24

 math

-0.24

å¼ķèµ·

-0.24

è²ī

-0.24

Eur

-0.24

ç¤¼

-0.23

POSITIVE LOGITS

æŀľ

0.29

åĩ¡æľ¬ç½ĳ

0.27

ÐµÐ½ÑĮ

0.26

éªĿ

0.25

beit

0.25

POOL

0.24

%{

0.24

pton

0.23

.gov

0.23

 ******************************************************************************/↵

0.23

Activations Density 0.046%

No Known Activations

This feature has no known activations.