INDEX

Explanations

code, technical terms, licensed text

The neuron detects C/C++ preprocessor directives, especially “#include <...>” statements.

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

 картофе

-0.87

зирован

-0.84

parr

-0.83

続

-0.83

мера

-0.81

the

-0.81

 gantung

-0.80

 KEYS

-0.79

deles

-0.79

biome

-0.79

POSITIVE LOGITS

 AttributeSet

1.07

ducción

1.07

 prover

1.02

 доказа

1.02

ABULARY

1.00

 witnesses

0.98

 witnessing

0.93

chartInstance

0.90

 Satisfaction

0.90

dom

0.90

Activations Density 0.024%