INDEX

Explanations

was

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_23.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.23.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 anon

-0.28

ä»ĬæĻļ

-0.27

åĩłå¼ł

-0.27

 fries

-0.26

wig

-0.26

å¹´ä¹ĭä¹ħ

-0.25

æĬ¤èº«

-0.25

-aged

-0.25

ä»Ĭå¤©çļĦ

-0.24

ä¸įåĲĥ

-0.24

POSITIVE LOGITS

origin

0.36

aim

0.35

 created

0.34

 originally

0.33

aim

0.32

 aiming

0.31

 aimed

0.30

æĺ¯ä¸ºäºĨ

0.30

 origin

0.28

Orig

0.28

Activations Density 0.024%

was

No Comments

No Known Activations