INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_20.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.20.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lrt

-0.32

allah

-0.28

 clubhouse

-0.26

 Brace

-0.26

laughter

-0.25

_:

-0.25

 Babe

-0.25

ddit

-0.25

 expos

-0.24

 expenditures

-0.24

POSITIVE LOGITS

oust

0.28

tract

0.28

eco

0.27

ec

0.27

æ¸¸

0.26

Select

0.26

ographics

0.26

umas

0.24

interop

0.24

misc

0.24

Activations Density 35.662%

No Known Activations

This feature has no known activations.