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

Negative Logits

åİŁæĿ¥æĺ¯

-0.27

paralle

-0.27

 gotta

-0.26

æīĢéľĢè¦ģçļĦ

-0.26

ä¸İæĹ¶ä¿±

-0.25

 improv

-0.25

 suce

-0.25

Spi

-0.25

æĽ´å¤ļç²¾å½©

-0.24

codile

-0.24

POSITIVE LOGITS

çĽĨ

0.30

ÐµÐ±

0.27

ä½¿åĳ½

0.26

éħ¸

0.25

(commands

0.24

åģļæ³ķ

0.23

åĽŀ

0.23

çļĦåģļæ³ķ

0.23

åħĪ

0.23

iani

0.23

Activations Density 3.764%

No Known Activations

This feature has no known activations.