INDEX

Explanations

key-value pairs

The neuron is looking for punctuation characters (especially quotation marks, periods, and commas) in code/JSON 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

濑

-1.14

al

-1.10

ne

-1.07

('

-1.05

AL

-1.05

for

-1.04

con

-1.04

su

-1.02

-1.00

POSITIVE LOGITS

今日も

1.37

 vôtre

1.35

先週

1.31

 veramente

1.30

 nastro

1.26

 comentários

1.26

前回の

1.25

Très

1.23

Интере

1.21

 diing

1.21

Activations Density 0.031%