INDEX

Explanations

track record

The neuron activates on mentions of a “track record” (i.e. references to past performance or records).

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

爌

-2.03

 gedeelte

-1.79

 forteller

-1.73

 voyez

-1.71

ᨗ

-1.65

quiao

-1.64

 besø

-1.63

 brukt

-1.63

蚺

-1.60

\{

-1.59

POSITIVE LOGITS

2.13

ut

1.96

os

1.94

ことも

1.91

as

1.87

fa

1.80

ta

1.78

 تضيف

1.77

1.74

 mere

1.73

Activations Density 0.002%