INDEX

Explanations

code syntax and descriptions

The neuron strongly activates on the “( … )” and trailing “;” patterns of function or constructor calls (e.g. this(...);, M(...), registerAllExtensions(...), InlineData(...)).

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

 into

-1.72

 where

-1.45

 Comprometido

-1.20

-1.19

for

-1.19

 appropriately

-1.14

由於

-1.13

 tabPage

-1.12

nintendo

-1.10

왠

-1.09

POSITIVE LOGITS

but

1.41

。“

1.33

[]:

1.29

nisch

1.27

 quien

1.27

 después

1.27

Someone

1.27

 sólo

1.26

cinogen

1.25

 alguien

1.24

Activations Density 0.018%