INDEX

Explanations

"makes", "defeats", "squadre", "drużyna", "takım"

This neuron detects numeric tokens formatted as 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

 продукт

0.40

 многие

0.39

ேத்க

0.38

DWARF

0.38

 выяв

0.38

Dieses

0.36

如果您

0.36

 имеется

0.36

 MANY

0.36

 πολλά

0.36

POSITIVE LOGITS

 yaşanan

0.44

 druż

0.43

 duelo

0.43

 murderous

0.43

 squadre

0.43

 bohater

0.42

 derrot

0.42

 ходе

0.40

 venge

0.40

 takım

0.40

Activations Density 0.001%