INDEX

Explanations

stores data

The neuron detects SQL clauses specifying storage format, in particular the “STORED AS TEXT FILE” phrase.

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

 кровь

-1.44

 reconnaît

-1.31

 ferait

-1.30

-1.23

-!

-1.22

∓

-1.21

 reçoit

-1.19

花了

-1.17

 MONEY

-1.17

 сво

-1.17

POSITIVE LOGITS

for

2.02

 storing

1.39

use

1.35

 storage

1.34

 menyimpan

1.29

 Imágenes

1.27

lidas

1.22

 using

1.21

 selanjutnya

1.20

 encanto

1.20

Activations Density 0.103%