INDEX

Explanations

defense systems and options

The neuron detects terms referring to military actions or operations (e.g. attack or defense).

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

</strong>

-1.98

ület

-1.80

What

-1.78

我一直

-1.76

𝕎

-1.74

導致

-1.72

 ſch

-1.68

ется

-1.65

Concerning

-1.63

 Eigentü

-1.63

POSITIVE LOGITS

呖

2.27

2.05

of

1.80

↵↵

1.69

媵

1.68

︲

1.67

 senza

1.61

🫅

1.61

闞

1.61

🇧

1.59

Activations Density 0.009%