INDEX

Explanations

to investigate, explore, examine

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

ativní

-0.89

 actrice

-0.85

μένο

-0.85

 hábit

-0.84

 jeito

-0.83

 impresion

-0.82

ListComponent

-0.81

reisen

-0.81

plained

-0.81

Geological

-0.79

POSITIVE LOGITS

 investigate

3.77

 explore

3.19

 investigating

3.05

 investigated

3.00

 investigates

2.97

 investigation

2.81

 examine

2.80

 explored

2.50

 بررسی

2.45

 exploring

2.44

Activations Density 0.503%