INDEX

Explanations

code syntax for assignment

This neuron detects Ruby-style named-parameter labels (hash keys) in code.

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

</h3>

-1.68

 rangement

-1.41

 menghadapi

-1.39

 servido

-1.38

部は

-1.34

 aparición

-1.34

ثيق

-1.34

--"

-1.34

 which

-1.31

hosting

-1.30

POSITIVE LOGITS

of

1.44

 другому

1.43

 في

1.41

喁

1.41

 журнала

1.40

全然

1.35

1.34

𓆏

1.34

}^{-}$

1.34

轰隆隆

1.33

Activations Density 0.004%