INDEX

Explanations

seeing and understanding things

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

 although

0.68

 perhaps

0.68

 ensuring

0.65

 resembles

0.64

 similarly

0.61

 resemble

0.61

 example

0.59

 nevertheless

0.59

有不少

0.59

 conjunction

0.59

POSITIVE LOGITS

 Your

2.13

 Them

1.92

 Their

1.85

Our

1.72

Your

1.71

 Yourself

1.65

 That

1.60

You

1.46

 Those

1.45

 Some

1.45

Activations Density 0.319%