INDEX

Explanations

file operations and path handling

The neuron peaks on standalone API, library, or framework identifiers (e.g. class names, module names, or function/namespace tokens) in code.

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

 were

-1.42

不断

-1.37

不要

-1.37

 quedado

-1.36

 with

-1.32

<>

-1.30

 weiße

-1.30

每年

-1.28

 vicenda

-1.26

鹇

-1.26

POSITIVE LOGITS

to

1.34

cestry

1.34

 dicha

1.33

 sabias

1.33

aknya

1.29

Pascal

1.28

atchewan

1.27

皱眉

1.26

袿

1.25

瞟

1.23

Activations Density 0.020%