INDEX

Explanations

description of following words

The neuron primarily fires on numeric tokens containing decimal points (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

 maneira

0.91

염

0.88

 членов

0.87

 maneiras

0.82

 король

0.81

 cenderung

0.80

 گردد

0.80

ற்றில்

0.78

 matematik

0.78

 पहला

0.77

POSITIVE LOGITS

INTE

0.87

turquoise

0.83

EXPL

0.83

IN

0.80

UN

0.77

UND

0.73

ADN

0.71

ACITY

0.71

INTER

0.68

GRU

0.68

Activations Density 0.001%