INDEX

Explanations

method/function calls

The neuron fires on code tokens that name or invoke functions/methods (i.e. identifiers used in calls, especially verbs like validate, update, draw, disable, appear, 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

 花纹

-1.35

 nimmer

-1.24

 Projektu

-1.20

 serien

-1.17

 ideen

-1.16

 solides

-1.15

鳚

-1.15

 permett

-1.15

 lachen

-1.14

蠵

-1.14

POSITIVE LOGITS

();

1.71

();

1.58

('

1.21

 will

1.16

 provide

1.16

().

1.14

 các

1.13

()));

1.11

All

1.11

("

1.10

Activations Density 0.009%