INDEX

Explanations

closing parentheses and braces

The neuron activates on conditional checks—especially comparisons against null or boolean values—in “if” 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

 koch

-1.41



-1.32



-1.30

 €.

-1.27

 dezelve

-1.27

formerly

-1.26

 eenig

-1.24

嫱

-1.23

 eenen

-1.23

⅜

-1.22

POSITIVE LOGITS

 with

1.52

地元

1.49

 Kemudian

1.42

Then

1.41

 this

1.29

スポンサーリンク

1.23

多くの

1.22

 जहां

1.20

You

1.16

 before

1.15

Activations Density 0.015%