INDEX

Explanations

merge or add items to collections

The neuron strongly activates on leading whitespace (indentation) in code blocks.

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

 deportivas

-0.81

 lectores

-0.79

veien

-0.76

URNS

-0.75

laughter

-0.75

Detta

-0.75

堟

-0.74

 dừng

-0.73

 целях

-0.73

lauch

-0.72

POSITIVE LOGITS

addAll

2.17

AddRange

1.70

putAll

1.57

extend

1.16

 extend

1.03

 combine

0.97

 combining

0.94

 merger

0.92

Combining

0.91

 merge

0.90

Activations Density 0.007%