INDEX

Explanations

rt

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_4.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.4.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

itus

-0.28

åĦ¿

-0.25

ãĤ¹ãĥĶãĥ¼ãĥī

-0.24

rido

-0.24

ventus

-0.24

latlong

-0.23

Ð´ÐµÐ½

-0.23

uttle

-0.23

inger

-0.23

itore

-0.23

POSITIVE LOGITS

 sword

0.28

ions

0.28

 exactly

0.28

unsafe

0.26

è´Ńè½¦

0.26

 striped

0.26

 Jordan

0.25

Ã¤ter

0.24

channel

0.24

é¾Ľ

0.24

Activations Density 0.020%

rt

No Comments

No Known Activations