INDEX

Explanations

XML namespaces and build scripts

The neuron strongly responds to programming and markup identifiers—especially all-caps acronyms or proper names in code or XML (e.g. CMAKE, System, Forms, flash, jetbrains).

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

 both

-1.19

 while

-1.16

 with

-1.07

 EVERY

-1.02

 every

-1.02

all

-0.99

 which

-0.97

 before

-0.96

 then

-0.95

Why

-0.95

POSITIVE LOGITS

 MATTERS

1.24

この

1.07

 brilhante

1.04

 trám

0.98

thered

0.98

 pageable

0.97

 diesem

0.97

 rzeczy

0.94

 इस

0.94

tyimages

0.94

Activations Density 0.071%