INDEX

Explanations

a bond

This neuron strongly activates on the indefinite articles “a” and “an.”

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

Popular

0.68

AT

0.65

OT

0.65

PLACEMENT

0.65

metaTag

0.65

ТЕ

0.64

Strategic

0.63

MBOL

0.63

定番

0.63

適切な

0.63

POSITIVE LOGITS

 phonons

0.64

 cosas

0.63

 ponctués

0.63

 carcinomas

0.63

 debacle

0.61

 nanotubes

0.60

 CMakeFiles

0.59

 cytokines

0.59

 cyclohex

0.58

 невозможно

0.58

Activations Density 0.003%