INDEX

Explanations

too early, too often, too tempting

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

veyard

2.01

</code>

1.86

ۡ

1.79

 Rows

1.78

 चयन

1.77

াবের

1.76

вода

1.74

似乎

1.72

いっぱい

1.72

Насе

1.71

POSITIVE LOGITS

$-(

2.27

ritional

2.21

ITED

2.04

$+

2.02

πηρε

2.00

issime

1.98

 gốc

1.98

issimi

1.94

睨

1.94

ited

1.92

Activations Density 0.058%