INDEX

Explanations

code delimiters

The main thing this neuron does is detect numeric literal tokens (especially floating-point constants) in source code.

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

 necessary

-1.37

 finland

-1.24

最近は

-1.23

 yüzük

-1.22

 these

-1.21

 berlina

-1.20

 veterinario

-1.20

Nazionale

-1.20

 Portugu

-1.20

↑↑↑

-1.20

POSITIVE LOGITS

 '*':

1.35

 szen

1.23

 whom

1.23

providedIn

1.20

 depois

1.20

gether

1.14

1.13

'',

1.13

 viel

1.13

 beide

1.13

Activations Density 0.056%