INDEX

Explanations

code identifiers and snippets

The neuron flags programming‐language tokens—especially low-level hardware/SIMD intrinsic names and numeric literal constants.

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

no

-1.38

-1.24

 voordat

-1.14

🌟

-1.14

 This

-1.14

-1.13

🔥🔥

-1.12

🔥🔥🔥

-1.10

Zunanje

-1.08

POSITIVE LOGITS

<bos>

1.57

そのため

1.48

ですが

1.40

がありますが

1.36

 ekst

1.33

は

1.30

 дори

1.29

 souverain

1.27

せっかく

1.27

même

1.26

Activations Density 0.002%