INDEX

Explanations

time durations

The neuron activates on expressions of temporal durations (e.g. “weeks,” “months,” “years,” “decades,” “time to come,” etc.).

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

 vrou

-1.48

 formel

-1.37

 citer

-1.37

 idéia

-1.34

 forcé

-1.32

 velha

-1.31

demokrat

-1.30

teken

-1.28

ھے

-1.22

 aussit

-1.19

POSITIVE LOGITS

we

1.59

 this

1.49

 then

1.45

 there

1.44

it

1.35

1.33

1.31

 things

1.30

1.25

1.20

Activations Density 0.038%