INDEX

Explanations

grade levels

The neuron fires on numeric grade or level indicators (e.g. “grade 3–5,” “levels 1990.2898,” “Academic level,” etc.).

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

bär

0.68

spec

0.59

Documentation

0.58

BBB

0.57

Playback

0.55

Containers

0.54

robots

0.54

MAX

0.54

BD

0.52

supports

0.52

POSITIVE LOGITS

 ideally

0.64

 prosecuted

0.59

 tractable

0.58

 guts

0.57

 आढ

0.56

 thirties

0.56

ਧੀ

0.56

 deceptive

0.55

 octagon

0.55

 willst

0.55

Activations Density 0.000%