INDEX

Explanations

court of appeals jurisprudence

The neuron activates on capitalized named entities—proper names of people, places, organizations, or specific addresses.

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

</h1>

-1.27

を行いました

-1.23

のもの

-1.23

に挑戦

-1.20

りました

-1.18

メス

-1.16

を試

-1.16

-1.14

 Coloque

-1.13

 защото

-1.12

POSITIVE LOGITS

توضیحات

1.34

Sinopse

1.31

middelen

1.30

Preparazione

1.30

TRIBUN

1.28

eradish

1.25

دانلود

1.20

تحميل

1.17

しかし

1.16

 לאחר

1.16

Activations Density 0.003%