INDEX

Explanations

code keywords and structures

The neuron is attuned to C/C++ source‐code constructs—especially preprocessor directives and header‐file tokens (e.g. “#include”, “#define”, “.h”).

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

”,

-1.39

rdoba

-1.27

“

-1.25

>=</

-1.25

antlr

-1.24

our

-1.24

cljs

-1.23

 бампер

-1.20

},

-1.19

 ayude

-1.17

POSITIVE LOGITS

想过

1.50

皈

1.34

 ſeveral

1.32

下载附件

1.25

 Kochi

1.23

惚

1.23

 rémun

1.22

 několika

1.20

 celebró

1.20

课堂

1.20

Activations Density 0.034%