INDEX

Explanations

instructions and tutorials

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_27.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.27.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 after

-0.39

after

-0.34

 dopo

-0.34

 aprÃ¨s

-0.31

 afterwards

-0.30

 afterward

-0.30

 Ð¿Ð¾ÑģÐ»Ðµ

-0.29

 suddenly

-0.29

After

-0.29

 immediately

-0.28

POSITIVE LOGITS

coat

0.28

åĪĿæŃ¥

0.26

ystone

0.26

ç®¡çĲĨå±Ģ

0.25

erk

0.24

éĢĤåºĶ

0.24

Moder

0.24

eding

0.23

guide

0.23

éĢĤ

0.23

Activations Density 0.003%

instructions and tutorials

No Known Activations