INDEX

Explanations

The neuron specializes in spotting sentence-initial discourse markers or transitional words that launch a new clause.

questions starting with Q

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

at

-1.23

my

-1.21

𝗺

-1.20

if

-1.10

for

-1.10

gården

-1.03

♥♥

-1.00

'),

-0.98

กว่า

-0.96

one

-0.95

POSITIVE LOGITS

 관련

1.34

 related

1.24

 liées

1.18

んだよね

1.18

 Vitamina

1.11

没什么

1.06

 ベージュ

1.04

 ケア

1.03

 liée

1.02

اااا

1.02

Activations Density 0.000%