INDEX

Explanations

he, she, names, followed by action or feeling

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Apparently

0.93

apparently

0.93

 apparently

0.89

 seeming

0.89

Apparently

0.87

 apparent

0.86

 meille

0.79

我们也

0.77

居然

0.76

 craziness

0.76

POSITIVE LOGITS

 knew

1.20

 feel

1.19

 feels

1.18

 watched

1.14

 knows

1.13

 জানে

1.11

feel

1.07

 imagines

1.05

感觉到

1.02

 siente

1.01

Activations Density 0.101%