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

èĢĺ

-0.27

 parti

-0.27

igrated

-0.26

estar

-0.26

ocha

-0.26

ASA

-0.26

ated

-0.25

nl

-0.25

osta

-0.25

geh

-0.25

POSITIVE LOGITS

aki

0.28

(Constant

0.27

çĢĽ

0.27

wife

0.27

bourne

0.26

ä¿Ĺ

0.26

å¹¸

0.24

well

0.24

ake

0.24

iam

0.24

Activations Density 0.002%

No Known Activations

This feature has no known activations.