INDEX

Explanations

Sheppard

The neuron is keyed to pick out tokens containing consecutive repeated letters (e.g. “weeping,” “seepage,” “Sheppard,” “Newsweek”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 apolog

-2.91

谞

-2.70

澐

-2.61

has

-2.53

 ensures

-2.50

 blatantly

-2.50

 为什么

-2.48

 maneuver

-2.38

𓍯

-2.38

罂

-2.34

POSITIVE LOGITS

 представителей

2.72

 stalwart

2.47

 MAJOR

2.42

2.25

噠

2.11

 البته

2.09

ians

2.09

pecabe

2.08

2.06

would

2.05

Activations Density 0.001%