INDEX

Explanations

use strict, literal

Neuron 4 fires on language-specific “strict mode” or literal‐freezing declarations (e.g. PHP’s `declare(strict_types=1)`, Ruby’s `# frozen_string_literal`, Python’s `from future import …`, Perl’s `use strict`, etc.).

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

 enough

-1.21

づけ

-1.16

 such

-1.12

の方に

-1.09

fch

-1.02

 utilizzati

-1.02

 interessanti

-1.02

 silêncio

-1.02

 espelho

-1.00

 pueden

-1.00

POSITIVE LOGITS

 waistcoat

1.13

蔵庫

1.09

<bos>

1.03

érios

1.02

但这

1.00

これも

0.97

bleau

0.95

fördert

0.93

 classiques

0.93

Nyc

0.93

Activations Density 0.002%