INDEX

Explanations

simple, low-effort, low-cost

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

substitution

1.06

favor

1.05

 MIUI

1.03

Salman

1.02

 preferring

1.00

Mozilla

1.00

chatbot

0.99

 لیون

0.99

despite

0.99

thanks

0.99

POSITIVE LOGITS

感覺

1.07

没什么

0.93

感觉

0.92

凄い

0.92

绝对

0.90

也不知道

0.89

没人

0.86

eway

0.85

充满了

0.83

强度

0.83

Activations Density 0.111%