INDEX

Explanations

code syntax and structure

The neuron activates on numeric values and integer type annotations (i.e. tokens representing numbers or the “int” keyword).

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

 tía

-1.48

 nép

-1.44

 bén

-1.43

ఌ

-1.38

 nagu

-1.38

 ekip

-1.37

 kimdir

-1.37

 TRABAJO

-1.36

他才

-1.35

\%,

-1.34

POSITIVE LOGITS

on

1.81

 with

1.63

由於

1.41

 while

1.38

länd

1.34

 "../../../../

1.33

These

1.32

 dicendo

1.32

這個

1.32

 незавер

1.30

Activations Density 0.004%