INDEX

Explanations

new date or object instantiation

The neuron fires on programming‐language reserved words and built‐in constants/functions (e.g. null/None, new, class, array, c, DateTime, 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

of

-1.93

◜

-1.22

 with

-1.21

嗨

-1.14

メルカリ

-1.13

-1.12

 esternos

-1.11

生活中

-1.10

『

-1.10

 focused

-1.09

POSITIVE LOGITS

oczes

1.47

 anledning

1.34

erweise

1.27

 Strukt

1.23

datei

1.21

ulose

1.19

 akku

1.17

ⓒ

1.16

 himo

1.15

 måte

1.15

Activations Density 0.095%