INDEX

Explanations

challenges and storms

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

𝑎

3.58

𝑜

3.53

𝑐

3.44

𝑢

3.27

𝑟

3.25

ו

3.17

𝑡

3.16

ⓒ

3.15

𝑠

3.12

𝑒

3.05

POSITIVE LOGITS

egraphics

3.32

ee

2.94

Појединачна

2.87

ead

2.79

ুমাত্র

2.78

evice

2.73

["

2.72

eeee

2.69

সঙ্গত

2.68

tedir

2.66

Activations Density 0.399%