INDEX

Explanations

Matrix, human, industry, cancer

The neuron selectively lights up on verbs and nouns in exhortative or self-improvement contexts—that is, action-oriented “calls to change” language (e.g. “continue,” “doing,” “hope,” “see,” “grow,” etc.).

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

 그의

0.77

 అతని

0.77

 inasmuch

0.77

 이용

0.73

 Tübingen

0.73

 elucidation

0.71

 accompan

0.70

 său

0.70

 treatment

0.69

 analogous

0.69

POSITIVE LOGITS

covid

1.23

 covid

1.21

🥳

1.19

 pandemia

1.16

 emojis

1.13

tiktok

1.13

 COVID

1.13

 Fortnite

1.11

Covid

1.10

 metaverse

1.09

Activations Density 0.232%