INDEX

Explanations

but then

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

 numerous

2.09

 extensively

2.09

 العديد

2.03

 many

2.03

 רבים

2.00

ingle

2.00

 utilizes

1.93

均为

1.91

 typically

1.89

 рекомендуется

1.86

POSITIVE LOGITS

 Anyway

2.28

Anyway

2.25

…?

2.01

...?

1.91

then

1.91

 پھر

1.90

怅

1.89

 THEN

1.88

Then

1.88

?..

1.86

Activations Density 0.123%