INDEX

Explanations

code constructs

This neuron detects C/C++ “struct” definitions and their field‐declaration lines.

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

 היתה

-1.30

̳

-1.20

 viena

-1.15

༎

-1.14

hetically

-1.11

 zatem

-1.10

 diante

-1.09

 somente

-1.09

 with

-1.09

 vienas

-1.08

POSITIVE LOGITS

 stagn

1.20

 MANY

1.13

互相

1.12

嫖

1.11

zeria

1.11

miendo

1.11

甚至是

1.10

 ESPEC

1.08

 однако

1.06

大体

1.05

Activations Density 0.121%