INDEX

Explanations

then preceding a past descriptor

The neuron selectively fires on the word “then” when it’s used to introduce a former title, role, or point in time.

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

騏

-1.29

 these

-1.27

 exces

-1.24

 CHECKBOX

-1.23

というわけで

-1.23

 MATHEMAT

-1.21

zadas

-1.20

 Infine

-1.20

 DICTION

-1.20

 coste

-1.18

POSITIVE LOGITS

1.40

1.34

By

1.34

 тогда

1.27

 alors

1.23

ing

1.23

Then

1.19

But

1.19

1.18

 használat

1.17

Activations Density 0.008%