INDEX

Explanations

dots, numbers, and symbols

The neuron activates primarily on numeric tokens with decimal points—that is, floating-point numbers.

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

 personen

0.61

 perust

0.55

eine

0.54

 osoby

0.54

编

0.52

 espí

0.52

 masyarakat

0.51

作家

0.50

 pracę

0.50

 uczest

0.50

POSITIVE LOGITS

 Theorems

0.68

 しかし

0.67

 Cnemidophorus

0.65

 pInBuffer

0.65

 Comparisons

0.65

 nField

0.62

 있지만

0.61

 Biochim

0.61

 ถึง

0.61

 CharPtr

0.61

Activations Density 0.000%