INDEX

Explanations

possessive followed by attribute

This neuron strongly activates on possessive pronouns (e.g., his, my, whose, your).

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

 newfound

0.98

 burgeoning

0.94

 многочис

0.86

 tumultuous

0.82

 upbringing

0.82

 increasingly

0.77

 myriad

0.77

 pervasive

0.76

诸多

0.75

 impending

0.73

POSITIVE LOGITS

 사람이

0.66

 целью

0.63

 문제는

0.62

目的是

0.61

 فقط

0.61

 Databases

0.61

 영역

0.59

 নিয়ম

0.58

GDP

0.56

 тільки

0.56

Activations Density 0.470%