INDEX

Explanations

programmatic intents and actions

The neuron fires on code identifiers (variable, field, and object names) in programming snippets.

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

 that

-1.35

 which

-1.34

 their

-1.30

any

-1.27

 this

-1.22

 after

-1.14

 whose

-1.13

the

-1.05

 what

-1.05

 whether

-1.02

POSITIVE LOGITS

...";

1.32

 monstru

1.32

↑↑↑

1.30

 alho

1.28

🤮

1.21

 marseille

1.20

 tulum

1.19

 HARVARD

1.19

😫

1.18

 haji

1.17

Activations Density 0.039%