INDEX

Explanations

code object or function definitions

This neuron activates on programming‐language keywords that name or invoke definition/registering functions or constructs (e.g. Register, defineMessages, extend, group, return).

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

 from

-1.23

 also

-1.21

 this

-1.19

-1.13

のお知らせ

-1.09

thodoxy

-1.06

сение

-1.05

ådet

-1.04

to

-1.03

POSITIVE LOGITS

<0xA3>

1.16

Bjs

1.16

 lesquelles

1.14

 весьма

1.13

ⴱ

1.12

junctive

1.11

 américa

1.10

ebenarnya

1.10

 lesquels

1.09

 galite

1.09

Activations Density 0.000%