INDEX

Explanations

math and logic questions

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 their

-1.02

 that

-1.00

 Toilette

-0.98

the

-0.98

 característica

-0.98

-0.94

urlopen

-0.94

ypos

-0.91

 totiž

-0.91

 sogenannte

-0.89

POSITIVE LOGITS

それとも

1.09

Kunden

1.09

Belum

1.06

或是

1.06

Paket

1.05

Kelas

1.02

بیشتر

1.02

Erste

1.02

两种

1.01

 vies

1.01

Activations Density 0.025%

No Known Activations