INDEX

Explanations

date and time

The neuron fires on mentions of “date” and especially “time” (often together as “Date/Time”) labels or fields.

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

 काफी

-1.90

現状

-1.55

inderella

-1.55

 яке

-1.52

At

-1.51

 only

-1.46

 кілька

-1.45

 οποία

-1.45

 from

-1.44

 яка

-1.43

POSITIVE LOGITS

of

2.16

沨

1.74

我国

1.74

ISCHE

1.73

ſche

1.71

閙

1.69

 diesem

1.68

戗

1.68

–)

1.64

镦

1.64

Activations Density 0.005%