INDEX

Explanations

hyphen

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_22.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.22.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 squeezing

-0.28

lin

-0.27

trg

-0.26

 tubing

-0.25

 squeezed

-0.25

egal

-0.25

ngr

-0.25

 objs

-0.24

.converter

-0.24

rtle

-0.23

POSITIVE LOGITS

åıĸ

0.30

 Take

0.30

Take

0.29

orns

0.29

éĹ´çļĦ

0.29

çĽ¸åħ³çļĦ

0.27

 take

0.27

 Takes

0.27

akes

0.26

ivas

0.25

Activations Density 0.032%

hyphen

No Known Activations