INDEX

Explanations

maximum threshold

The neuron responds to code‐like comment or directive lines—especially ones beginning with “#” (e.g. “# Set the comparison criteria”)—i.e. it flags comment headers or configuration directives.

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

computed

0.61

 해당

0.59

Qualified

0.59

옛

0.59

Abstract

0.58

NEXT

0.57

 Zustand

0.57

旧

0.56

처

0.55

듯

0.55

POSITIVE LOGITS

 threshold

1.24

 tolerance

1.16

 maximum

1.15

 tolerancia

1.15

 maksimum

1.13

 timeout

1.10

 radius

1.08

threshold

1.08

max

1.07

maxlen

1.05

Activations Density 0.110%