INDEX

Explanations

plural nouns and their attributes

The neuron fires on plural noun tokens, i.e. words ending in “s” that name resources or entities (e.g. users, cars, groups).

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

tanleria

-1.60

 этого

-1.57

ܗ

-1.52

 offroad

-1.48

浤

-1.46

ｚ

-1.41

 ПОД

-1.39

ギュア

-1.38

出一

-1.36

CIÓN

-1.35

POSITIVE LOGITS

</strong>

2.08

1.70

–

1.45

 semua

1.38

1.35

all

1.33

 выполнения

1.29

who

1.25

him

1.22

Activations Density 0.240%