INDEX

Explanations

closing code blocks

This neuron fires on procedure‐call expressions—i.e. identifiers immediately followed by “(…)” such as system(...), printInstructions(), getUserAction(), etc.

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

on

-1.65

in

-1.54

out

-1.51

In

-1.43

-1.40

so

-1.38

IN

-1.32

-1.30

-1.28

-1.27

POSITIVE LOGITS

 then

2.42

 followed

1.56

 następnie

1.52

 rồi

1.52

然後

1.51

 سپس

1.50

แล้ว

1.49

 ثم

1.48

 vervolgens

1.48

 Followed

1.38

Activations Density 0.030%