INDEX

Explanations

activities and actions

The neuron is keyed to multiword prepositional or connective phrases (e.g. “in response to,” “through the day,” “in terms of,” “planned for this”) that link clauses or express relationships.

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

 possibility

-1.54

for

-1.32

 bicara

-1.27

 khả

-1.22

 ability

-1.22

优秀的

-1.16

↓↓↓

-1.15

 Since

-1.15

Արտաքին

-1.15

has

-1.12

POSITIVE LOGITS

 including

1.58

↵↵↵

1.34

 such

1.30

 SUCH

1.28

 toward

1.28

Including

1.27

↵↵↵↵↵↵↵

1.26

↵↵↵↵↵↵

1.22

nin

1.20

↵↵↵↵↵↵↵↵↵

1.20

Activations Density 0.256%