INDEX

Explanations

is a definition

The neuron detects definitional or explanatory phrases—places where a term is introduced and defined (e.g., “X is …,” “X are …,” or “X is defined as …”).

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

 with

0.87

 including

0.66

with

0.62

 excluding

0.61

(!)

0.61

had

0.60

 both

0.60

wt

0.60

 without

0.59

and

0.58

POSITIVE LOGITS

是一種

0.80

 понятие

0.79

 refiere

0.78

一種

0.75

 जेव्हा

0.72

 असते

0.71

 процесс

0.71

现象

0.71

現象

0.71

と呼ばれる

0.71

Activations Density 0.123%