INDEX

Explanations

property, places, and titles

The neuron strongly responds to numeric tokens—quantities, measurements, dates, and other numbers in the text.

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

 Herald

-0.90

gpa

-0.82

ित्र

-0.81

CUC

-0.79

Fuzzy

-0.77

 上午

-0.77

大会

-0.77

 Lumpur

-0.75

 reloading

-0.75

achten

-0.75

POSITIVE LOGITS

 flats

0.79

 residents

0.77

Residents

0.72

 housing

0.72

 type

0.71

 Types

0.69

 водо

0.69

 ocze

0.69

 types

0.68

 ин

0.68

Activations Density 0.010%