INDEX

Explanations

problems, issues, errors

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

soluble

1.89

 destinados

1.87

瞰

1.85

 subordination

1.82

 ましょう

1.80

義務

1.77

Hed

1.77

financed

1.76

supervised

1.74

という意味

1.73

POSITIVE LOGITS

 sorun

1.83

 persiste

1.75

 problème

1.66

 problemi

1.60

 problema

1.58

 problèmes

1.54

 ocorre

1.50

 sporad

1.49

 stutter

1.49

 symptômes

1.49

Activations Density 0.353%