INDEX

Explanations

ed. or eds. followed by names

The neuron detects short bibliographic role abbreviations (e.g. “Ed.”, “ed.”, “Arr.”) marking editors or arrangers in reference entries.

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



-1.62

 matières

-1.62

撄

-1.59

怍

-1.57

 condominio

-1.55

 quantités

-1.51

 preventivo

-1.49

 étan

-1.49

 credito

-1.48

 ſec

-1.47

POSITIVE LOGITS

by

2.86

 with

2.13

 from

2.08

 more

1.91

for

1.87

at

1.85

all

1.78

 primarily

1.73

.”

1.66

Activations Density 0.046%