INDEX

Explanations

strategies leading to aid

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

Qar

1.44

 Bogor

1.43

 Toro

1.34

PSB

1.28

QP

1.27

 Sait

1.26

 Circuit

1.26

+'

1.25

ติ

1.25

 Xiong

1.25

POSITIVE LOGITS

Kelly

1.57

 Kelly

1.36

 kelly

1.17

Plate

1.14

 Sometimes

1.13

有时

1.13

Sometimes

1.13

Occasionally

1.04

kelly

1.04

autog

1.01

Activations Density 0.067%

No Known Activations