INDEX

Explanations

reptiles and amphibians

This neuron fires on any occurrence of “turtle” (including “turtles”), whether in the context of the Python turtle graphics library or references to actual turtles.

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

 ಕಬ್ಬಿಣ

0.63

 Albright

0.62

 Bethlehem

0.56

Containers

0.54

 casks

0.53

沖

0.52

containers

0.52

Cun

0.52

 bateau

0.52

 Jericho

0.51

POSITIVE LOGITS

 turtle

1.44

 turtles

1.40

 Turtles

1.39

🐢

1.34

 Turtle

1.33

Turtle

1.30

turtle

1.23

 reptile

1.10

龟

1.07

龜

1.05

Activations Density 0.109%