INDEX

Explanations

andinonat

The neuron primarily detects very common short function words (e.g. “and,” “in,” “is,” “a”).

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

Our

-1.19

 Hinsicht

-1.13

 tohoto

-1.12

 této

-1.11

 квітня

-1.09

 مكتبه

-1.09

 olacak

-1.09

我们的

-1.07

 otwor

-1.06

 individuellen

-1.04

POSITIVE LOGITS

😁

1.54

😊

1.50

🙂

1.48

VGND

1.38

😀

1.36

🙄

1.36

👍👍

1.35

 🤩

1.35

cloths

1.34

romeda

1.34

Activations Density 0.195%