INDEX

Explanations

temporal references

This neuron responds to temporal adverbial phrases indicating when something happened (e.g. “the other day,” “last week,” “in November,” “recently”).

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

 they

-1.63

 despite

-1.41

 Despite

-1.41

ким

-1.40

 there

-1.40

Sonstiges

-1.39

we

-1.38

也太

-1.36

 mockery

-1.35

 assaulting

-1.33

POSITIVE LOGITS

绷

1.41

 uniquement

1.35

 encuent

1.34

 occidentalis

1.33



1.29

 favoriser

1.27

 mycket

1.25

見えます

1.24

なども

1.22

ходите

1.22

Activations Density 0.106%