INDEX

Explanations

what follows i, you, they

The neuron selectively activates on numeric tokens that include decimal points (floating-point numbers).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ی

1.52

ます

1.35

י

1.24

た

1.19

る

1.13

ee

1.13

бва

1.12

 prostatic

1.12

ில்

1.11

ي

1.08

POSITIVE LOGITS

 وما

1.29

 касается

1.14

 transpired

1.10

 alumno

1.03

 appunto

1.01

देशीर

0.98

 nume

0.97

 préc

0.97

 يلي

0.97

 phép

0.96

Activations Density 0.153%