INDEX

Explanations

known for

The neuron activates on occurrences of the word “considered” (as in “is/are considered to be/as”) indicating evaluative or descriptive phrases.

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

 nondegenerate

0.79

১

0.77

ENTION

0.75

 WRITE

0.75

 APPEAL

0.73

みた

0.73

含义

0.73

 посредством

0.72

 Сколько

0.72

৪

0.72

POSITIVE LOGITS

oks

0.93

ably

0.85

្នក

0.83

ość

0.79

 damals

0.79

unek

0.79

 दर्जा

0.79

stoffen

0.75

fares

0.75

esthetics

0.75

Activations Density 0.045%