INDEX

Explanations

copyright notices

The neuron detects four-digit year numbers (e.g., “1999,” “2001,” “2003”) typically used in copyright or date stamps.

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

就没有

-1.28

边的

-1.25

 propietarios

-1.21

当然是

-1.20

よる

-1.19

assem

-1.16

没问题

-1.15

 coreana

-1.14

 jejich

-1.13

 Their

-1.12

POSITIVE LOGITS

to

1.25

翠

1.23

↵↵↵

1.17

mathrm

1.15

 ammonia

1.10

徘

1.05

 stof

1.05

klere

1.05

 besonder

1.03

Specifies

1.02

Activations Density 0.288%