INDEX

Explanations

web links or URLs

The neuron detects parts of web addresses (URL components such as “www” or “https”).

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

 personalización

-1.30

जय

-1.22

钾

-1.20

 Nedir

-1.20

忘记

-1.20

 Añade

-1.18

 كأس

-1.16

maxn

-1.15

 oameni

-1.14

 basadas

-1.09

POSITIVE LOGITS

at

1.52

 will

1.30

for

1.26

 with

1.14

and

1.13

TRIBUN

1.12

by

1.11

or

1.09

on

1.08

Frage

1.04

Activations Density 0.000%