INDEX

Explanations

number or issue

The neuron activates on “No” or “Number” (and variants like “n”) immediately followed by digits, i.e. references to issue/publication numbers.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 that

-2.09

and

-1.35

 from

-1.34

 allow

-1.28

返品

-1.23

 necessary

-1.22

ResumeLayout

-1.21

any

-1.20

 even

-1.19

 mitään

-1.19

POSITIVE LOGITS

Cevap

1.16



1.12

Воспитатель

1.10

Shaft

1.05

Răsp

1.04

嘮

1.03

corazón

1.03

があるので

1.03

 många

1.02

 verschillende

1.02

Activations Density 0.003%