INDEX

Explanations

sports team abbreviations and names

The neuron flags rare or low‐frequency subword tokens—e.g. obscure proper names, technical terms, or unusual character sequences.

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

-2.08

 which

-1.90

 when

-1.88

on

-1.77

 that

-1.72

if

-1.70

 just

-1.62

is

-1.58

-1.42

↵

-1.41

POSITIVE LOGITS

 damski

1.78

 phénom

1.68

岽

1.64

ými

1.63

carus

1.59

zzet

1.57

 italiens

1.55

 ventanas

1.53

\}}

1.51

doesn

1.50

Activations Density 0.141%