INDEX

Explanations

checking if something

This neuron responds to occurrences of checking logic, especially the word “Check” (often in comments) and “if” conditional statements.

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

<bos>

-1.55

ativen

-1.27

WITHOUT

-1.23

hence

-1.20

hasta

-1.17

igné

-1.17

知道的

-1.16

 reconocidos

-1.16

を利用

-1.14

<0xA3>

-1.13

POSITIVE LOGITS

 whether

2.86

if

2.22

 that

2.05

how

1.80

 against

1.59

 which

1.45

是否

1.43

for

1.42

 apakah

1.37

 where

1.34

Activations Density 0.029%