INDEX

Explanations

code linting or style checking

The neuron activates on mentions of code‐linting tools or lint directives (e.g. “lint,” “pylint,” “eslint,” etc.).

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

することは

-1.41

Surprisingly

-1.23

Though

-1.23

疟

-1.21

轰

-1.20

epiece

-1.19

olyte

-1.18

 Unlike

-1.17

ubre

-1.17

of

-1.16

POSITIVE LOGITS

 then

1.55

 belangrijkste

1.52

still

1.42

 trochę

1.32

already

1.23

 algemene

1.17

ícias

1.16

is

1.15

also

1.13

 bijzondere

1.12

Activations Density 0.130%