INDEX

Explanations

adjectival endings followed by nouns

The neuron reliably activates on Slavic adjectival or participial forms ending in the suffix “ного” (e.g. нормального, машинного, произвольного).

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

-2.89

and

-2.59

↵↵

-2.38

-2.27

are

-2.19

-2.16

感叹

-2.13

 plunging

-2.00

 proverbial

-1.97

廚

-1.92

POSITIVE LOGITS

новь

2.69

”—

2.67

酲

2.64

亊

2.59

ઁ

2.58

стреча

2.55

獃

2.50

’，

2.47

 повіт

2.45

спомина

2.42

Activations Density 0.004%