INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_22.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.22.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

.samples

-0.27

orda

-0.26

æ¼«

-0.26

çıŀ

-0.25

è¾°

-0.25

ç¼ģ

-0.24

æİĴ

-0.24

-common

-0.24

itter

-0.24

åħ±

-0.24

POSITIVE LOGITS

 Turnbull

0.32

vic

0.27

atical

0.26

nist

0.26

 plain

0.25

CSR

0.24

anmar

0.23

ä¸Ĭæľī

0.23

|-

0.23

 ballpark

0.23

Activations Density 0.000%

No Known Activations

This feature has no known activations.