INDEX

Explanations

code punctuation and symbols

This neuron fires on numeric literals (and their surrounding punctuation), e.g. numbers in code, version strings, file‐path indexes, or other digit sequences.

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

を選ぶ

-1.39

MAXN

-1.37

 your

-1.29

的で

-1.28

 kompet

-1.27

aguay

-1.19

鹇

-1.17

 asiático

-1.15

他说

-1.13

 previamente

-1.11

POSITIVE LOGITS

↵↵

1.27

↵

1.25

 will

1.24

 käns

1.20

 båda

1.13

 möjlighet

1.10

 смотреть

1.10

 của

1.08

 ибо

1.06

!!!!

1.05

Activations Density 0.098%