INDEX

Explanations

types of form

The neuron activates on words that label categories or classifications—e.g. “type,” “form,” “kind,” or “forms.”

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

 parachoque

-2.22

ación

-1.95

penuhi

-1.93

there

-1.92

 $^{\

-1.91

 tussen

-1.85

-1.79

却不

-1.78

 recognizes

-1.77

打量

-1.74

POSITIVE LOGITS

2.09

“

2.05

 form

1.91

1.89

 skriva

1.84

.“

1.83

“

1.79

。“

1.77

ANIEL

1.76

 включая

1.75

Activations Density 0.056%