INDEX

Explanations

news headlines, documentation

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

èĮĥåĽ´

-0.30

æĹĮ

-0.27

æĲª

-0.27

çļĦæĸ¹åĲĳ

-0.26

Domain

-0.25

Ð²Ñĥ

-0.25

hil

-0.25

ObjectName

-0.25

åĩłä½ķ

-0.25

æī£

-0.24

POSITIVE LOGITS

ssp

0.27

bara

0.26

 sideways

0.25

iana

0.24

 parody

0.24

èĭ¥æĤ¨

0.24

 Kramer

0.24

 disjoint

0.24

 spaghetti

0.24

SSA

0.24

Activations Density 0.001%

news headlines, documentation

No Known Activations