INDEX

Explanations

tying, knots, and bows

The main thing this neuron does is find words related to tying or fastening (e.g., tie, tied, knot, knotted).

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

vailable

-1.48

er

-1.35

Thru

-1.27

 ----

-1.24

oneof

-1.23

 ------

-1.22

while

-1.21

tempt

-1.20

 سنة

-1.20

身體

-1.20

POSITIVE LOGITS

夀

1.35

 tying

1.34

 Bolshe

1.33

 당신

1.32

もら

1.31

':

1.31

ษ

1.31

 оригі

1.30

pijama

1.29

ے

1.28

Activations Density 0.033%