INDEX

Explanations

after conjunctions and punctuation

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

No Comments

Negative Logits

.scalablytyped

-0.10

 majority

-0.10

åĺī

-0.09

 )\n\n\n\n\n\n\n\n

-0.09

 weiber

-0.09

trak

-0.09

RuntimeObject

-0.09

Ä±ÅŁÄ±k

-0.09

pane

-0.09

ÐĲÑĢÑħÑĸÐ²

-0.09

POSITIVE LOGITS

 there

0.12

it

0.11

 plenty

0.09

 modulo

0.09

 lord

0.09

any

0.09

 Given

0.09

 suffice

0.09

 given

0.09

 again

0.09

Activations Density 0.020%