INDEX

Explanations

mathematical operations

The neuron fires on numeric tokens (digits/sequences of digits), spotting numbers in the text.

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

 edge

0.61

 hennes

0.61

她是

0.59

 EDGE

0.58

 inodes

0.57

MIP

0.57

 edged

0.56

 OEMs

0.56

 vrchol

0.56

 inches

0.55

POSITIVE LOGITS

Math

0.62

0.59

Free

0.59

Safe

0.58

Fluid

0.58

わり

0.57

_=

0.57

sinh

0.57

Reducing

0.56

ua

0.55

Activations Density 0.001%