INDEX

Explanations

always followed by a state or action

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

 example

1.46

 allemaal

1.44

に関する法律

1.42

每个

1.40

all

1.40

所有人

1.38

 dérou

1.35

 entire

1.34

 Typical

1.33

大きい

1.33

POSITIVE LOGITS

cky

2.14

ymers

2.11

orre

2.08

greens

2.02

caution

1.95

perp

1.93

_+

1.93

ếp

1.92

clus

1.88

plr

1.87

Activations Density 0.156%