INDEX

Explanations

pile of

The neuron detects mentions of things being gathered or heaped into piles or piles-like accumulations.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

↵↵

-0.98

}+

-0.81

ルフ

-0.80

 bỏ

-0.80

 процедуры

-0.79

-0.78

 пла

-0.77

 wild

-0.77

 horrible

-0.77

POSITIVE LOGITS

driver

1.03

 cadáver

0.97

heap

0.96

Výhody

0.91

 cómics

0.90

 DRIVE

0.89

Heap

0.87

 jste

0.87

 aimé

0.86

")[

0.86

Activations Density 0.014%