INDEX

Explanations

witch

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

çĮ´

-0.28

çĶ²

-0.28

peg

-0.27

 åħ

-0.26

åĲĵ

-0.24

 mate

-0.24

å½±åĵį

-0.24

Dx

-0.23

ä¼ĺ

-0.23

å¸¦èµ°

-0.23

POSITIVE LOGITS

furt

0.30

iskey

0.26

å¯¿åĳ½

0.26

qualities

0.26

åĬĽãĤĴ

0.25

å£½

0.25

iy

0.25

vol

0.25

ouce

0.25

Assoc

0.24

Activations Density 0.002%

witch

No Comments

No Known Activations