INDEX

Explanations

conjunctions and prepositions

The neuron detects prominent descriptive words—especially adjectives and adverbs highlighting features or qualities (e.g. “covered,” “fenced,” “landscaped,” “secure,” “environmentally,” “powerful,” etc.).

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

its

-1.48

 verschiedenen

-1.33

前进

-1.26

>(),

-1.26

 sentada

-1.25

那样

-1.25

emon

-1.23

 conjunta

-1.23

ніше

-1.23

翆

-1.22

POSITIVE LOGITS

 from

1.52

 called

1.42

 looking

1.35

 they

1.28

 provato

1.27

for

1.26

 equipped

1.25

to

1.23

 needing

1.23

at

1.21

Activations Density 0.117%