INDEX

Explanations

This neuron detects the names of fields, tags, or attributes (the keys) in structured data formats like XML, JSON, or config files.

structured data fields and tags

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

for

-1.29

 this

-1.25

 from

-1.25

any

-1.21

if

-1.20

and

-1.09

 that

-1.06

in

-1.06

 プロフィール

-1.04

to

-1.03

POSITIVE LOGITS

えません

1.18

 aceptas

1.16

selben

1.13

aliya

1.13



1.09

zel

1.08

ויים

1.08

 одновременно

1.05

pn

1.03

 acepción

1.03

Activations Density 0.001%