INDEX

Explanations

paying attention to

This neuron detects instances of directing focus or notice—especially the phrase “attention to” (and variants like “paying attention to”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

"};

-1.15

事儿

-1.13

derdag

-1.07

极致

-1.06

 évaluation

-1.05

ezt

-1.05

ilfe

-1.04

接种

-1.04

culable

-1.04

,’

-1.03

POSITIVE LOGITS

to

1.72

how

1.65

the

1.44

which

1.38

 where

1.31

ஂ

1.28

 важно

1.27

菝

1.19

 intently

1.19

 perché

1.19

Activations Density 0.013%