INDEX

Explanations

you are / it is

This neuron detects direct second-person address—especially “you” (often in “If you…” conditional phrases).

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

probably

1.09

Probably

1.08

almost

1.02

 probablement

1.00

 probabilmente

0.95

 Probably

0.95

 probably

0.93

 майже

0.93

 تقریباً

0.92

Almost

0.90

POSITIVE LOGITS

 хотите

1.29

 चाहें

1.22

 want

1.17

 ever

1.16

 quiser

1.16

 souhaitez

1.15

 herhangi

1.15

 असेल

1.09

 dudas

1.08

 chcesz

1.08

Activations Density 1.191%