INDEX

Explanations

preparation and purpose

The neuron responds to procedural or instructional content—text describing actions, steps, or technical processes.

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

枚目

-1.16

 needing

-0.94

みなさま

-0.94

 exclusivo

-0.93

 飲み

-0.91

 होना

-0.91

 Olimpia

-0.90

 utilizzo

-0.90

途径

-0.90

 tork

-0.89

POSITIVE LOGITS

to

1.66

and

1.38

for

1.35

เพื่อ

1.35

 чтобы

1.20

 để

1.17

 برای

1.13

就可以

1.06

 afin

1.03

 щоб

1.02

Activations Density 0.110%