INDEX

Explanations

nonprofit organization or Delaware corporation

The main thing this neuron does is find sequences of digits (numeric tokens) in the text.

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

athed

-1.26

 techniques

-1.11

 município

-1.09

 plenty

-1.09

知っている

-1.05

 グラン

-1.04

abinieri

-1.03

 kinds

-1.01

可能

-1.01

 possible

-1.01

POSITIVE LOGITS

OGRAPH

1.31

ּוֹ

1.27

rophi

1.24

<bos>

1.22

 originalmente

1.21

ulando

1.17

ֵּ

1.15

ñores

1.13

ādā

1.12

пів

1.12

Activations Density 0.041%