INDEX

Explanations

bow vs, brush, NDC, Siena, set

This neuron activates on uncommon multi‐syllable nouns—especially proper names or specialized technical terms (e.g. brand names, people’s names, place names, or domain-specific jargon).

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

it

-2.34

 albeit

-1.93

have

-1.93

and

-1.92

since

-1.86

if

-1.80

 является

-1.71

even

-1.70

-1.69

some

-1.67

POSITIVE LOGITS

。

1.91

！"

1.76

 pienso

1.75

橚

1.73



1.73

"){

1.73

踅

1.71

 メニュー

1.70

efits

1.67

 такого

1.66

Activations Density 0.041%