INDEX

Explanations

some topics and examples

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

isible

2.64

್

2.47

м

2.39

zano

2.18

iał

2.16

ському

2.15

aż

2.14

 било

2.12

夊

2.10

ihe

2.09

POSITIVE LOGITS

`;

2.27

WORD

2.15

ones

2.04

offs

2.02

<$

2.00

ok

1.96

psia

1.95

remove

1.91

here

1.91

ਹਿ

1.90

Activations Density 0.316%