INDEX

Explanations

definitely followed by action verb

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

黍

2.97

ಿಕೆ

2.86

zok

2.70

嵋

2.67

THING

2.67

 мах

2.64

днев

2.54

ង្វ

2.53

сть

2.52

<bos>

2.49

POSITIVE LOGITS

waves

2.61

uin

2.30

aries

2.18

inc

2.16

nelly

2.14

برا

2.11

nement

2.07

arses

2.06

2.05

ality

2.04

Activations Density 0.025%