INDEX

Explanations

code or question syntax

This neuron detects the literal field or label “description.”

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

 will

-1.05

has

-1.04

an

-1.02

had

-0.97

 untuk

-0.91

-0.90

tom

-0.90

—

-0.90

-0.87

you

-0.87

POSITIVE LOGITS

Keterangan

1.34

 тонн

1.22

 Boletín

1.16

 setelan

1.15

1.14

 Apalagi

1.13

齟

1.13

媖

1.11

améliorer

1.10

 Berikut

1.10

Activations Density 0.000%