INDEX

Explanations

operator sequence

The neuron fires on code‐navigation metadata lines—things like “Definition at line X of file Y,” symbol references, and similar file/line markers.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

lũ

0.59

__(

0.54

reinterpret

0.52

grand

0.52

 characterised

0.52

:],

0.52

 stdout

0.50

 static

0.49

ilient

0.49

 confirmer

0.49

POSITIVE LOGITS

operator

1.04

 operator

1.02

 operators

0.88

 Operators

0.88

运营商

0.85

 Operator

0.82

Operator

0.80

operators

0.80

Operators

0.79

emplace

0.77

Activations Density 0.014%