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

rame

-0.29

hurst

-0.25

uggy

-0.24

äººèĦī

-0.24

ogie

-0.24

èĩªå·±åĸľæ¬¢

-0.24

åİĨæĹ¶

-0.24

ä¸ĭè½¦

-0.23

ipline

-0.23

 aggregates

-0.23

POSITIVE LOGITS

Serialized

0.28

 Serial

0.27

Nou

0.26

 SERIAL

0.26

Serial

0.26

çķ²

0.25

 serial

0.25

_serial

0.25

 foreground

0.24

ot

0.24

Activations Density 0.197%

No Known Activations

This feature has no known activations.