INDEX

Explanations

closed doors, public sessions

The neuron detects words that mark formal procedural or legal actions and settings (e.g. dismissals, rulings, removals, open/public sessions in courts or assemblies).

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

of

-1.51

 Também

-1.25

 επίσης

-1.24

 want

-1.23

 need

-1.23

 mudanças

-1.20

 reveal

-1.15

 také

-1.14

șa

-1.14

또

-1.12

POSITIVE LOGITS

小弟

1.28

├

1.23

 prenez

1.23

нтів

1.22

uding

1.20

 BERNAMA

1.19

biendo

1.16

wicklung

1.16

），

1.16

 GALERÍA

1.16

Activations Density 0.019%