INDEX

Explanations

This neuron detects identifiers (variable or object names) in code.

programming variable access

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

if

-1.47

 menikmati

-1.27

yoruz

-1.16

توضیحات

-1.16

 that

-1.16

 with

-1.14

and

-1.10

 their

-1.10

Роди

-1.05

suje

-1.05

POSITIVE LOGITS

 didn

1.22

 hadn

1.21

 wasn

1.13

1.09

is

0.99

ମ

0.98

 Lösungs

0.96

 weren

0.95

 fatis

0.94

 couldn

0.93

Activations Density 0.003%