INDEX

Explanations

software, accounts, and networks

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_25.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.25.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

åĪ¶åº¦

-0.41

ç®¡çĲĨåĪ¶åº¦

-0.31

 system

-0.30

alone

-0.28

çĶ¨å¾®ä¿¡

-0.27

åĿĩ

-0.27

 Mitt

-0.27

ç³»ç»Ł

-0.26

ç®¡çĲĨä½ĵç³»

-0.26

 mechanism

-0.26

POSITIVE LOGITS

(&_

0.30

icro

0.25

åħ¶å®ĥéĹ®é¢ĺ

0.24

nier

0.24

 Corrections

0.24

lers

0.23

 Hundred

0.23

spb

0.23

,&

0.23

rowsable

0.23

Activations Density 0.104%

software, accounts, and networks

No Comments

No Known Activations