INDEX

Explanations

code delimiters and assignments

This neuron responds to code-specific tokens—especially punctuation like semicolons (;) and commas (,) and numeric literals—i.e. it detects punctuation marks and numbers 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

 both

-2.22

//

-2.06

="

-1.90

by

-1.88

 first

-1.84

Both

-1.84

re

-1.80

two

-1.65

 putting

-1.63

 creating

-1.61

POSITIVE LOGITS

 帅气

1.90



1.88

gangssprach

1.78

盦

1.78

 insanely

1.77

 potrebné

1.77

 ส์

1.74

demás

1.73

the

1.73

⮑

1.70

Activations Density 0.018%