INDEX

Explanations

past intentions

This neuron detects expressions of future intent or planned actions (e.g. “was going to,” “going to,” “planning to”).

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

 pierwszej

-1.23

有時

-1.23

 will

-1.17

不时

-1.17

 vannak

-1.09

たまに

-1.08

 måde

-1.05

切った

-1.05

 vacía

-1.02

有时候

-1.01

POSITIVE LOGITS

but

1.82

 tetapi

1.38

to

1.36

原本

1.34

 originally

1.32

 mutta

1.21

 وكان

1.20

 yesterday

1.19

if

1.16

 ancak

1.14

Activations Density 0.028%