INDEX

Explanations

enough to enable outcomes

This neuron fires on occurrences of the word “enough,” especially when used as an intensifier (often in “enough to…” constructions).

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

-1.90

INS

-1.64

 колба

-1.58

 immense

-1.55

 spécial

-1.48

 extremely

-1.48

 sowieso

-1.47

нії

-1.44

蠵

-1.43

 телефона

-1.41

POSITIVE LOGITS

 would

1.78

してみました

1.69

 ''){

1.60

 started

1.59

出發

1.58

 tells

1.57

 potrafi

1.53

۱۸

1.53

ستم

1.51

esistenza

1.51

Activations Density 0.007%