INDEX

Explanations

electrical devices, materials, and phenomena

The neuron selectively responds to occurrences of the word “electrical” (and its close morphological variants).

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



-2.31

 эти

-2.27

琿

-2.20

-2.19

cherichia

-2.11

来た

-2.06

缡

-2.03

HostException

-2.03

pinephrine

-2.02

ガール

-1.98

POSITIVE LOGITS

</h3>

3.36

3.05

quele

2.92

 беже

2.86

al

2.83

 your

2.78

ၟ

2.78



2.73

by

2.66

摃

2.66

Activations Density 0.012%