INDEX

Explanations

RipQ Co

This neuron detects tokens that serve as code‐declaration identifiers—e.g. language keywords, type or annotation names, and access modifiers in programming source.

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

＼(^

-1.02

！(

-1.01

Украина

-0.99

(´

-0.97

лк

-0.96

 without

-0.93

توضیحات

-0.92

ﻖ

-0.90

ENTINA

-0.90

ñar

-0.88

POSITIVE LOGITS

0.95

JPanel

0.93

 panel

0.93

 JPanel

0.92

May

0.83

只有一

0.83

 panneaux

0.81

Activations Density 0.000%