INDEX

Explanations

for followed by entity

The neuron strongly fires on instances of the preposition “for” when it introduces a numeric specification (e.g. “for 180,” “for 33,” “for 151,” 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

 harapan

-1.51

⃢

-1.40

Mə

-1.36

 история

-1.33

あった

-1.33

Ferdin

-1.31

 duvar

-1.30

想い

-1.30

 ortam

-1.30

muito

-1.30

POSITIVE LOGITS

1.48

for

1.39

1.28

entierung

1.25

’

1.25

vae

1.23

ってきて

1.23

1.22

uelles

1.16

Activations Density 0.102%