INDEX

Explanations

The neuron detects common sentence-initial discourse markers or subordinating conjunctions (e.g. “With,” “As,” “On,” “After,” “In”) that introduce clauses.

As, With, after

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

but

-1.60

 with

-1.48

 only

-1.45

as

-1.35

 perhaps

-1.29

 which

-1.28

for

-1.22

 both

-1.19

in

-1.13

on

-1.11

POSITIVE LOGITS

 groet

1.03

ところに

1.02

 درباره

0.98

 professione

0.97

凄く

0.93

 király

0.92

conclusion

0.91

AUGUST

0.91

 sooo

0.91

 aéri

0.89

Activations Density 0.040%