INDEX

Explanations

testing code and specific queries

The neuron is primarily triggered by Java class or library identifiers (e.g. “Test” and “tika”).

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

ș

-1.77

 encantó

-1.55

taya

-1.42

</h1>

-1.40

antism

-1.40

thodoxy

-1.39

abytes

-1.36

了好

-1.35

thodox

-1.35

 encantan

-1.34

POSITIVE LOGITS

 ritard

1.41

</em>

1.30

 bunga

1.27

 goreng

1.26

 before

1.24

تأ

1.23

葙

1.21

 dollari

1.19

to

1.18

 beurre

1.17

Activations Density 0.000%