INDEX

Explanations

XML and configuration fragments

This neuron flags data‐type tag names (e.g. “string,” “real,” “array”) in structured/markup documents.

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

by

-1.73

and

-1.56

 just

-1.51

on

-1.39

 created

-1.38

for

-1.38

In

-1.33

 designed

-1.33

 over

-1.31

 found

-1.30

POSITIVE LOGITS

Alamat

1.52

 tropicales

1.38

Belum

1.38

 garder

1.34

 rempla

1.34

 inond

1.33

 matchs

1.33

Pilih

1.32

льних

1.31

 orquesta

1.30

Activations Density 0.002%