INDEX

Explanations

let me introduce an action or explanation

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ederal

-0.90

 negozio

-0.89

 Synd

-0.88

 ragazzo

-0.88

 selam

-0.88

 konfig

-0.88

 maksimum

-0.87

ystema

-0.86

érées

-0.86

ïc

-0.85

POSITIVE LOGITS

be

1.58

 know

1.45

 tell

1.32

 help

1.09

 first

1.08

 explain

1.08

ราบ

1.05

 please

1.01

dieran

0.96

 сейчас

0.96

Activations Density 0.012%