INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_30.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.30.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

addy

-0.30

 Flags

-0.28

 Muse

-0.27

 Flag

-0.26

illary

-0.25

pdf

-0.24

Gus

-0.24

 viewType

-0.23

-uri

-0.23

phot

-0.23

POSITIVE LOGITS

tera

0.29

æ¼«éķ¿çļĦ

0.26

eking

0.25

INDER

0.25

çĽ¸ç»ĵåĲĪ

0.24

ëĭī

0.24

å°ģéĹŃ

0.24

dirname

0.24

 tight

0.24

raw

0.23

Activations Density 0.000%

No Known Activations

This feature has no known activations.