INDEX

Explanations

lists of content types

This neuron specializes in spotting plural noun phrases (multi‐word or single‐word) such as “slide shows,” “Twitter Chats,” “viewing parties,” “seed rounds,” “landing pages,” etc.

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

 appellee

0.46

 равна

0.46

 diffusivity

0.45

 ہے۔

0.43

 deaminase

0.42

 trừ

0.41

 должна

0.41

 sería

0.41

___

0.41

 reductase

0.41

POSITIVE LOGITS

 webinars

0.53

featuring

0.49

pdfs

0.48

 ebooks

0.48

creens

0.47

filenames

0.47

 PDFs

0.46

 galore

0.46

 videolar

0.46

大型

0.46

Activations Density 0.000%