INDEX

Explanations

single, different, every, other, same

The neuron is primarily sensitive to degree- and quantity-expressing words—comparatives, superlatives, and general quantifiers such as “more,” “most,” “many,” “other,” “any,” “better,” and “small.”

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

RELATIVA

0.53

RECTION

0.52

ง

0.52

 ยก

0.52

 Плавание

0.50

 絶対

0.50

MktCap

0.50

 아버지

0.50

 зокрема

0.49

 บ่

0.49

POSITIVE LOGITS

 companies

0.58

 countries

0.57

種類の

0.54

 precautions

0.52

 accessories

0.51

 soorten

0.51

 shrubs

0.50

 products

0.49

 firms

0.48

 manufacturers

0.48

Activations Density 0.727%