INDEX

Explanations

hexadecimal numbers and codes

The neuron is tuned to spot hexadecimal numeric literals (e.g. byte‐value tokens like “0xE8”, “xE8”, lone hex digits/bytes).

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

-2.06

-2.02

 ofrece

-1.98

-1.90

-1.88

 nuevos

-1.84

-1.82

 tendrá

-1.78

POSITIVE LOGITS

ghalaya

1.77

 izman

1.57

 }}$.

1.52

we

1.51

})=\

1.46

種類の

1.44

 jenem

1.43

($__

1.43

he

1.42

strator

1.42

Activations Density 0.006%