INDEX

Explanations

The neuron primarily activates on numeric sequences (digits) in bibliographic metadata such as volume, issue, page numbers, years, and DOIs.

fraction or numerical ranges

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

If

-0.96

 även

-0.86

蹴

-0.85

 rangkaian

-0.84

if

-0.84

If

-0.83

As

-0.82

as

-0.82

[]:

-0.81

acetate

-0.81

POSITIVE LOGITS

 Collateral

0.88

ܬ

0.86

 tortas

0.85

 Rücksicht

0.82

Répondre

0.82

exposition

0.82

Belgian

0.80

 yến

0.79

implies

0.79

 größe

0.78

Activations Density 0.010%