INDEX

Explanations

Here is an example

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

 energized

1.17

 invigorating

1.11

 stately

1.10

 delightfully

1.08

 resiliency

1.07

 commemorative

1.07

 directorial

1.07

 回転

1.06

 カバー

1.05

 leisurely

1.05

POSITIVE LOGITS

 اريد

1.22

Unable

1.17

 समस्या

1.16

 проблема

1.14

 хочу

1.14

 вопрос

1.09

problem

1.08

ปัญหา

1.08

困

1.07

Activations Density 0.466%