INDEX

Explanations

Brown

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_12.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.12.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

äºĨä¸ª

-0.29

ä¸ªåŁİå¸Ĥ

-0.29

istics

-0.28

Codes

-0.26

tek

-0.26

etic

-0.26

ARD

-0.25

 atomic

-0.25

åĪĩæĪĲ

-0.25

atin

-0.24

POSITIVE LOGITS

riangle

0.26

 Emil

0.25

ä¹īåĬ¡æķĻèĤ²

0.25

ç²Ł

0.24

 POSSIBILITY

0.24

respond

0.24

\f

0.23

çģ¯åħī

0.23

erti

0.23

-actions

0.23

Activations Density 0.001%

Brown

No Known Activations