INDEX

Explanations

stopwords or section

The neuron strongly activates on numeric tokens—especially multi‐part numbers or version‐style decimals.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

↵

0.51

0.46

льбе

0.44

 insanların

0.40

ريقيا

0.39

েও

0.39

ko

0.38

ire

0.38

 complementary

0.38

conoc

0.37

POSITIVE LOGITS

 للخ

0.49

ņas

0.46

 fuer

0.44

以後

0.44

 ढक

0.43

শকার

0.43

 под

0.43

しばらく

0.42

甘

0.42

 réaliser

0.42

Activations Density 0.000%