INDEX

Explanations

code snippets and keywords

The neuron detects numeric literals (number tokens) in code snippets.

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

to

-1.48

 only

-1.35

on

-1.30

 including

-1.26

 other

-1.21

all

-1.17

etc

-1.16

อื่น

-1.14

riebene

-1.10

 although

-1.05

POSITIVE LOGITS

if

1.25

;\\

1.22

で見る

1.20

さぁ

1.14

を行い

1.13

...".

1.11

 vervolgens

1.09

 тогда

1.05

 então

1.04

Activations Density 0.025%