INDEX

Explanations

see below

This neuron detects directional cues or calls to action pointing readers toward supplemental content (e.g. “see below,” “check out…video below,” “fill out the form below”).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ഗ്ര

0.53

рал

0.51

 નીક

0.51

 그거

0.51

 अवे

0.50

posizione

0.50

 फिल

0.50

 copertura

0.49

觡

0.49

 Sicily

0.49

POSITIVE LOGITS

 below

4.75

below

4.31

Below

4.03

 Below

4.00

 abaixo

3.81

下面

3.73

 ниже

3.69

 BELOW

3.64

 아래

3.56

下面的

3.55

Activations Density 0.733%