INDEX

Explanations

after event, outcome

The neuron fires on punctuation‐marked clause or sentence transitions—i.e. commas or sentence boundaries and the very next words (especially discourse markers like “After,” “I,” “the,” etc.).

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

 while

-1.55

 from

-1.41

 despite

-1.20

 although

-1.20

now

-1.13

 currently

-1.13

 meanwhile

-1.12

 today

-1.07

 σήμερα

-1.05

 until

-1.04

POSITIVE LOGITS

 وبعد

1.16

WARDS

1.00

 dasselbe

0.99

剩下的

0.97

 anschließend

0.94

retour

0.93

 tilbage

0.93

ayakan

0.92

atória

0.91

变得

0.89

Activations Density 0.070%