INDEX

Explanations

This neuron fires whenever the text launches into a first‐person, diary‐style anecdote—that is, the opening personal‐experience sentences of a blog post.

end of clause/sentence

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

you

-1.73

currently

-1.55

 today

-1.45

are

-1.35

 currently

-1.33

 your

-1.28

 tomorrow

-1.23

 сьогодні

-1.22

 here

-1.22

now

-1.20

POSITIVE LOGITS

was

2.34

 decided

1.84

 وكان

1.71

 knew

1.70

 было

1.53

 थे

1.45

だったので

1.42

 وكانت

1.41

 began

1.39

 seemed

1.37

Activations Density 0.058%