INDEX

Explanations

preposition "by" in many languages

The neuron responds to salient content words—relatively infrequent nouns and verbs that carry the key informational “weight” of a sentence.

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

никах

0.54

GLOBALS

0.53

ிலான

0.52

лед

0.52

ದಲ

0.51

ífico

0.50

 निर्माण

0.49

idences

0.49

বিজ

0.49

 куда

0.48

POSITIVE LOGITS

by

3.91

 oleh

3.36

 توسط

3.00

โดย

2.94

by

2.81

によって

2.72

 द्वारा

2.61

 tarafından

2.55

 โดย

2.52

 দ্বারা

2.50

Activations Density 1.390%