INDEX

Explanations

sentence ending punctuation

periods followed by the start of a new sentence, especially in dialogue or narrative contexts.

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 arguably

-0.10

 question

-0.10

 probably

-0.08

oco

-0.08

reen

-0.08

âīł

-0.08

ê¼

-0.08

ted

-0.08

ãģĿãĤĵãģª

-0.08

éĤ£ä¹Ī

-0.08

POSITIVE LOGITS

 apparently

0.41

 Apparently

0.38

Apparently

0.35

arently

0.26

 according

0.23

æį®

0.23

 According

0.21

æĵļ

0.20

according

0.20

According

0.18

Activations Density 0.125%