INDEX

Explanations

However, given, considering

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

または

1.14

もしくは

1.14

 oder

1.11

 hoặc

1.10

 অনুসারে

1.08

 alebo

1.06

ili

1.04

或者

1.03

或

1.01

 หรือ

1.01

POSITIVE LOGITS

Yet

1.35

なのに

1.31

 অথচ

1.30

Yet

1.24

yet

1.22

しかも

1.20

卻

1.17

 especially

1.17

especially

1.11

let

1.10

Activations Density 0.132%