INDEX

Explanations

possessive pronouns followed by nouns

The neuron specifically detects numeric tokens that include a decimal point (i.e. floating-point numbers).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 only

-2.08

for

-1.41

not

-1.40

是小

-1.38

 zobac

-1.36

 stolz

-1.33

 тільки

-1.31

 recibiendo

-1.30

 where

-1.30

 what

-1.28

POSITIVE LOGITS

骨折

1.45

portál

1.43

1.42

系列

1.40

”？

1.39

Fprintf

1.38

塒

1.37

💃

1.36

 serien

1.35

 regeringen

1.34

Activations Density 0.058%