INDEX

Explanations

close contests and victories

The neuron fires on numeric tokens and adjoining quantitative/statistical details (e.g. game scores, rounds, distances) in sports contexts.

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

 ทั้ง

-1.32

 knew

-1.30

 know

-1.29

 numerous

-1.26

忍者

-1.20

anjutan

-1.20

 Both

-1.16

 various

-1.16

一族

-1.16

л

-1.15

POSITIVE LOGITS

1.51

ers

1.48

if

1.32

amek

1.31

three

1.27

four

1.27

st

1.27

"];

1.24

ish

1.24

ro

1.24

Activations Density 0.024%