INDEX

Explanations

[@... ] citations

The neuron is triggered by numeric tokens, i.e., it detects numbers (especially multi‐digit identifiers or citations) in the text.

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

 Ráp

-1.24

 аккура

-1.21

 často

-1.16

 Índice

-1.16

gör

-1.14

laşı

-1.13

 Características

-1.13

 použití

-1.13

tarta

-1.13

Jangan

-1.11

POSITIVE LOGITS

1.38

one

1.16

1.15

no

1.14

1.12

1.09

1.08

et

1.04

…

1.02

Activations Density 0.040%