INDEX

Explanations

conjunction then word

The neuron lights up on content‐bearing “open‐class” words—nouns, verbs, adjectives, and adverbs—rather than on common function words.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 不能

0.88

 неправи

0.86

 zuerst

0.86

ไม่ได้

0.85

 nahin

0.83

Dijstra

0.80

 поэтому

0.78

ErrorClazz

0.78

 deswegen

0.77

 Jugendlichen

0.77

POSITIVE LOGITS

 your

0.81

 youll

0.77

you

0.75

↵

0.73

your

0.70

はもちろん

0.67

 możesz

0.66

 می‌توانید

0.65

You

0.64

 сможете

0.63

Activations Density 0.325%