INDEX

Explanations

what is happening or has happened

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

 ולא

1.32

 khá

1.30

 meski

1.28

 довольно

1.27

 chứ

1.26

 bukanlah

1.26

 કંઈ

1.25

 hefty

1.24

 Tidak

1.23

 Meski

1.22

POSITIVE LOGITS

 dihasilkan

1.18

 entailed

1.14

ervi

1.12

 bestowed

1.11

 underlies

1.06

 의한

1.05

 underlie

1.04

导致

1.03

φ

1.02

 dimaksud

1.02

Activations Density 0.641%