INDEX

Explanations

closing brackets and parentheses

This neuron fires on square‐bracket delimiters (the “[`” and “]” tokens) that mark array or list literals 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

was

-1.63

mirante

-1.41

ipfel

-1.40

簟

-1.39



-1.34

bB

-1.30

🎃

-1.29

ριο

-1.29

LITE

-1.28

ūs

-1.27

POSITIVE LOGITS

 that

1.93

’

1.45

1.38

 però

1.34

 podía

1.32

ください

1.32

,;

1.31

for

1.29

 however

1.28

 höheren

1.28

Activations Density 0.032%