INDEX

Explanations

here and now

The neuron is keyed to adverbs or adverbial prepositions that denote general or inclusive location (e.g. “here,” “everywhere,” “including”).

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

といいます

-1.16

らしいです

-1.10

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.02

 alcune

-1.01

 psicología

-1.00

and

-1.00

 ordine

-0.98

 soste

-0.97

 animale

-0.97

CONCLUS

-0.97

POSITIVE LOGITS

 when

1.40

ggy

1.34

in

1.27

ridescent

1.15

»>

1.10

nadequate

1.07

zzet

1.05

piracy

1.05

napp

1.03

 ",";

1.03

Activations Density 0.026%