INDEX

Explanations

programming and multi-language code snippets

text written in non-Latin scripts, particularly Hebrew, Cyrillic, Chinese, and Arabic characters.

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

jogo

-2.38

 Some

-2.34

 murió

-2.27

 Those

-2.20

な感じ

-2.19

 наша

-2.14

ędzynarod

-2.14

↵↵

-2.13

 dahulu

-2.11

 Interestingly

-2.09

POSITIVE LOGITS

3.48

3.27

3.08

2.95

2.83

2.69

co

2.63

sh

2.61

2.58

2.53

Activations Density 0.001%